Natural Language Processing: An Overview

Syracuse University
SURFACE
The School of Information Studies Faculty
Scholarship
School of Information Studies (iSchool)
2001
Natural Language Processing
Elizabeth D. Liddy
Syracuse University, [email protected]
Follow this and additional works at:
http://surface.syr.edu/istpub
Part of the Library and Information Science Commons, and the
Linguistics Commons
This Book Chapter is brought to you for free and open access by
the School of Information Studies (iSchool) at SURFACE. It has
been accepted for
inclusion in The School of Information Studies Faculty
Scholarship by an authorized administrator of SURFACE. For
more information, please contact
[email protected]
Recommended Citation
Liddy, E.D. 2001. Natural Language Processing. In
Encyclopedia of Library and Information Science, 2nd Ed. NY.
Marcel Decker, Inc.
http://surface.syr.edu?utm_source=surface.syr.edu%2Fistpub%2
F63&utm_medium=PDF&utm_campaign=PDFCoverPages

http://surface.syr.edu/istpub?utm_source=surface.syr.edu%2Fist
pub%2F63&utm_medium=PDF&utm_campaign=PDFCoverPage
s
s
http://surface.syr.edu/ischool?utm_source=surface.syr.edu%2Fis
tpub%2F63&utm_medium=PDF&utm_campaign=PDFCoverPage
s
s
http://network.bepress.com/hgg/discipline/1018?utm_source=sur
face.syr.edu%2Fistpub%2F63&utm_medium=PDF&utm_campai
gn=PDFCoverPages
http://network.bepress.com/hgg/discipline/371?utm_source=surf
ace.syr.edu%2Fistpub%2F63&utm_medium=PDF&utm_campaig
n=PDFCoverPages
mailto:[email protected]
Natural Language Processing
1
INTRODUCTION
Natural Language Processing (NLP) is the computerized
approach to analyzing text that
is based on both a set of theories and a set of technologies. And,
being a very active area
of research and development, there is not a single agreed-upon

definition that would
satisfy everyone, but there are some aspects, which would be
part of any knowledgeable
person’s definition. The definition I offer is:
Definition: Natural Language Processing is a theoretically
motivated range of
computational techniques for analyzing and representing
naturally occurring texts
at one or more levels of linguistic analysis for the purpose of
achieving human-like
language processing for a range of tasks or applications.
Several elements of this definition can be further detailed.
Firstly the imprecise notion of
‘range of computational techniques’ is necessary because there
are multiple methods or
techniques from which to choose to accomplish a particular type
of language analysis.
‘Naturally occurring texts’ can be of any language, mode,
genre, etc. The texts can be
oral or written. The only requirement is that they be in a

language used by humans to
communicate to one another. Also, the text being analyzed
should not be specifically
constructed for the purpose of the analysis, but rather that the
text be gathered from actual
usage.
The notion of ‘levels of linguistic analysis’ (to be further
explained in Section 2) refers to
the fact that there are multiple types of language processing
known to be at work when
humans produce or comprehend language. It is thought that
humans normally utilize all
of these levels since each level conveys different types of
meaning. But various NLP
systems utilize different levels, or combinations of levels of
linguistic analysis, and this is
seen in the differences amongst various NLP applications. This
also leads to much
confusion on the part of non-specialists as to what NLP really
is, because a system that
uses any subset of these levels of analysis can be said to be an
NLP-based system. The
difference between them, therefore, may actually be whether the

system uses ‘weak’ NLP
or ‘strong’ NLP.
‘Human-like language processing’ reveals that NLP is
considered a discipline within
Artificial Intelligence (AI). And while the full lineage of NLP
does depend on a number
of other disciplines, since NLP strives for human-like
performance, it is appropriate to
consider it an AI discipline.
‘For a range of tasks or applications’ points out that NLP is not
usually considered a
goal in and of itself, except perhaps for AI researchers. For
others, NLP is the means for
1
Liddy, E. D. In Encyclopedia of Library and Information
Science, 2
nd
Ed. Marcel Decker, Inc.
accomplishing a particular task. Therefore, you have

Information Retrieval (IR) systems
that utilize NLP, as well as Machine Translation (MT),
Question-Answering, etc.
Goal
The goal of NLP as stated above is “to accomplish human-like
language processing”.
The choice of the word ‘processing’ is very deliberate, and
should not be replaced with
‘understanding’. For although the field of NLP was originally
referred to as Natural
Language Understanding (NLU) in the early days of AI, it is
well agreed today that while
the goal of NLP is true NLU, that goal has not yet been
accomplished. A full NLU
System would be able to:
1. Paraphrase an input text
2. Translate the text into another language
3. Answer questions about the contents of the text
4. Draw inferences from the text

While NLP has made serious inroads into accomplishing goals 1
to 3, the fact that NLP
systems cannot, of themselves, draw inferences from text, NLU
still remains the goal of
NLP.
There are more practical goals for NLP, many related to the
particular application for
which it is being utilized. For example, an NLP-based IR system
has the goal of
providing more precise, complete information in response to a
user’s real information
need. The goal of the NLP system here is to represent the true
meaning and intent of the
user’s query, which can be expressed as naturally in everyday
language as if they were
speaking to a reference librarian. Also, the contents of the
documents that are being
searched will be represented at all their levels of meaning so
that a true match between
need and response can be found, no matter how either are
expressed in their surface form.
Origins

As most modern disciplines, the lineage of NLP is indeed
mixed, and still today has
strong emphases by different groups whose backgrounds are
more influenced by one or
another of the disciplines. Key among the contributors to the
discipline and practice of
NLP are: Linguistics - focuses on formal, structural models of
language and the
discovery of language universals - in fact the field of NLP was
originally referred to as
Computational Linguistics; Computer Science - is concerned
with developing internal
representations of data and efficient processing of these
structures, and; Cognitive
Psychology - looks at language usage as a window into human
cognitive processes, and
has the goal of modeling the use of language in a
psychologically plausible way.
Divisions
While the entire field is referred to as Natural Language
Processing, there are in fact two
distinct focuses – language processing and language generation.

The first of these refers
to the analysis of language for the purpose of producing a
meaningful representation,
while the latter refers to the production of language from a
representation. The task of
Natural Language Processing is equivalent to the role of
reader/listener, while the task of
Natural Language Generation is that of the writer/speaker.
While much of the theory and
technology are shared by these two divisions, Natural Language
Generation also requires
a planning capability. That is, the generation system requires a
plan or model of the goal
of the interaction in order to decide what the system should
generate at each point in an
interaction. We will focus on the task of natural language
analysis, as this is most
relevant to Library and Information Science.
Another distinction is traditionally made between language
understanding and speech
understanding. Speech understanding starts with, and speech
generation ends with, oral

language and therefore rely on the additional fields of acoustics
and phonology. Speech
understanding focuses on how the ‘sounds’ of language as
picked up by the system in the
form of acoustical waves are transcribed into recognizable
morphemes and words. Once
in this form, the same levels of processing which are utilized on
written text are utilized.
All of these levels, including the phonology level, will be
covered in Section 2; however,
the emphasis throughout will be on language in the written
form.
BRIEF HISTORY OF NATURAL LANGUAGE PROCESSING
Research in natural language processing has been going on for
several decades dating
back to the late 1940s. Machine translation (MT) was the first
computer-based
application related to natural language. While Weaver and
Booth (1); (2) started one of
the earliest MT projects in 1946 on computer translation based
on expertise in breaking

enemy codes during World War II, it was generally agreed that
it was Weaver’s
memorandum of 1949 that brought the idea of MT to general
notice and inspired many
projects (3). He suggested using ideas from cryptography and
information theory for
language translation. Research began at various research
institutions in the United States
within a few years.
Early work in MT took the simplistic view that the only
differences between languages
resided in their vocabularies and the permitted word orders.
Systems developed from this
perspective simply used dictionary-lookup for appropriate
words for translation and
reordered the words after translation to fit the word-order rules
of the target language,
without taking into account the lexical ambiguity inherent in
natural language. This
produced poor results. The apparent failure made researchers
realize that the task was a
lot harder than anticipated, and they needed a more adequate
theory of language.

However, it was not until 1957 when Chomsky (4) published
Syntactic Structures
introducing the idea of generative grammar, did the field gain
better insight into whether
or how mainstream linguistics could help MT.
During this period, other NLP application areas began to
emerge, such as speech
recognition. The language processing community and the speech
community then was
split into two camps with the language processing community
dominated by the
theoretical perspective of generative grammar and hostile to
statistical methods, and the
speech community dominated by statistical information theory
(5) and hostile to
theoretical linguistics (6).
Due to the developments of the syntactic theory of language and
parsing algorithms, there
was over-enthusiasm in the 1950s that people believed that fully
automatic high quality
translation systems (2) would be able to produce results
indistinguishable from those of

human translators, and such systems should be in operation
within a few years. It was not
only unrealistic given the then-available linguistic knowledge
and computer systems, but
also impossible in principle (3).
The inadequacies of then-existing systems, and perhaps
accompanied by the over-
enthusiasm, led to the ALPAC (Automatic Language Processing
Advisory Committee of
the National Academy of Science - National Research Council)
report of 1966. (7) The
report concluded that MT was not immediately achievable and
recommended it not be
funded. This had the effect of halting MT and most work in
other applications of NLP at
least within the United States.
Although there was a substantial decrease in NLP work during
the years after the ALPAC
report, there were some significant developments, both in
theoretical issues and in
construction of prototype systems. Theoretical work in the late
1960’s and early 1970’s

focused on the issue of how to represent meaning and
developing computationally
tractable solutions that the then-existing theories of grammar
were not able to produce. In
1965, Chomsky (8) introduced the transformational model of
linguistic competence.
However, the transformational generative grammars were too
syntactically oriented to
allow for semantic concerns. They also did not lend themselves
easily to computational
implementation. As a reaction to Chomsky’s theories and the
work of other
transformational generativists, case grammar of Fillmore, (9),
semantic networks of
Quillian, (10), and conceptual dependency theory of Schank,
(11) were developed to
explain syntactic anomalies, and provide semantic
representations. Augmented transition
networks of Woods, (12) extended the power of phrase-structure
grammar by
incorporating mechanisms from programming languages such as
LISP. Other
representation formalisms included Wilks’ preference semantics
(13), and Kay’s

functional grammar (14).
Alongside theoretical development, many prototype systems
were developed to
demonstrate the effectiveness of particular principles.
Weizenbaum’s ELIZA (15) was
built to replicate the conversation between a psychologist and a
patient, simply by
permuting or echoing the user input. Winograd’s SHRDLU (16)
simulated a robot that
manipulated blocks on a tabletop. Despite its limitations, it
showed that natural language
understanding was indeed possible for the computer (17).
PARRY (18) attempted to
embody a theory of paranoia in a system. Instead of single
keywords, it used groups of
keywords, and used synonyms if keywords were not found.
LUNAR was developed by
Woods (19) as an interface system to a database that consisted
of information about lunar
rock samples using augmented transition network and
procedural semantics (20).

In the late 1970’s, attention shifted to semantic issues,
discourse phenomena, and
communicative goals and plans (21). Grosz (22) analyzed task-
oriented dialogues and
proposed a theory to partition the discourse into units based on
her findings about the
relation between the structure of a task and the structure of the
task-oriented dialogue.
Mann and Thompson (23) developed Rhetorical Structure
Theory, attributing hierarchical
structure to discourse. Other researchers have also made
significant contributions,
including Hobbs and Rosenschein (24), Polanyi and Scha (25),
and Reichman (26).
This period also saw considerable work on natural language
generation. McKeown’s
discourse planner TEXT (27) and McDonald’s response
generator MUMMBLE (28) used
rhetorical predicates to produce declarative descriptions in the
form of short texts, usually
paragraphs. TEXT’s ability to generate coherent responses
online was considered a major
achievement.

In the early 1980s, motivated by the availability of critical
computational resources, the
growing awareness within each community of the limitations of
isolated solutions to NLP
problems (21), and a general push toward applications that
worked with language in a
broad, real-world context (6), researchers started re-examining
non-symbolic approaches
that had lost popularity in early days. By the end of 1980s,
symbolic approaches had been
used to address many significant problems in NLP and
statistical approaches were shown
to be complementary in many respects to symbolic approaches
(21).
In the last ten years of the millennium, the field was growing
rapidly. This can be
attributed to: a) increased availability of large amounts of
electronic text; b) availability
of computers with increased speed and memory; and c) the
advent of the Internet.
Statistical approaches succeeded in dealing with many generic
problems in computational
linguistics such as part-of-speech identification, word sense
disambiguation, etc., and

have become standard throughout NLP (29). NLP researchers
are now developing next
generation NLP systems that deal reasonably well with general
text and account for a
good portion of the variability and ambiguity of language.
LEVELS OF NATURAL LANGUAGE PROCESSING
The most explanatory method for presenting what actually
happens within a Natural
Language Processing system is by means of the ‘levels of
language’ approach. This is
also referred to as the synchronic model of language and is
distinguished from the earlier
sequential model, which hypothesizes that the levels of human
language processing
follow one another in a strictly sequential manner.
Psycholinguistic research suggests that
language processing is much more dynamic, as the levels can
interact in a variety of
orders. Introspection reveals that we frequently use information
we gain from what is
typically thought of as a higher level of processing to assist in a

lower level of analysis.
For example, the pragmatic knowledge that the document you
are reading is about
biology will be used when a particular word that has several
possible senses (or
meanings) is encountered, and the word will be interpreted as
having the biology sense.
Of necessity, the following description of levels will be
presented sequentially. The key
point here is that meaning is conveyed by each and every level
of language and that since
humans have been shown to use all levels of language to gain
understanding, the more
capable an NLP system is, the more levels of language it will
utilize.
(Figure 1: Synchronized Model of Language Processing)
Phonology
This level deals with the interpretation of speech sounds within
and across words. There

are, in fact, three types of rules used in phonological analysis:
1) phonetic rules – for
sounds within words; 2) phonemic rules – for variations of
pronunciation when words
are spoken together, and; 3) prosodic rules – for fluctuation in
stress and intonation
across a sentence. In an NLP system that accepts spoken input,
the sound waves are
analyzed and encoded into a digitized signal for interpretation
by various rules or by
comparison to the particular language model being utilized.
Morphology
This level deals with the componential nature of words, which
are composed of
morphemes – the smallest units of meaning. For example, the
word preregistration can
be morphologically analyzed into three separate morphemes: the
prefix pre, the root
registra, and the suffix tion. Since the meaning of each
morpheme remains the same
across words, humans can break down an unknown word into its
constituent morphemes

in order to understand its meaning. Similarly, an NLP system
can recognize the meaning
conveyed by each morpheme in order to gain and represent
meaning. For example,
adding the suffix –ed to a verb, conveys that the action of the
verb took place in the past.
This is a key piece of meaning, and in fact, is frequently only
evidenced in a text by the
use of the -ed morpheme.
Lexical
At this level, humans, as well as NLP systems, interpret the
meaning of individual words.
Several types of processing contribute to word-level
understanding – the first of these
being assignment of a single part-of-speech tag to each word. In
this processing, words
that can function as more than one part-of-speech are assigned
the most probable part-of-
speech tag based on the context in which they occur.
Additionally at the lexical level, those words that have only one
possible sense or

meaning can be replaced by a semantic representation of that
meaning. The nature of the
representation varies according to the semantic theory utilized
in the NLP system. The
following representation of the meaning of the word launch is in
the form of logical
predicates. As can be observed, a single lexical unit is
decomposed into its more basic
properties. Given that there is a set of semantic primitives used
across all words, these
simplified lexical representations make it possible to unify
meaning across words and to
produce complex interpretations, much the same as humans do.
launch (a large boat used for carrying people on rivers, lakes
harbors, etc.)
((CLASS BOAT) (PROPERTIES (LARGE)
(PURPOSE (PREDICATION (CLASS CARRY) (OBJECT
PEOPLE))))
The lexical level may require a lexicon, and the particular
approach taken by an NLP

system will determine whether a lexicon will be utilized, as well
as the nature and extent
of information that is encoded in the lexicon. Lexicons may be
quite simple, with only
the words and their part(s)-of-speech, or may be increasingly
complex and contain
information on the semantic class of the word, what arguments
it takes, and the semantic
limitations on these arguments, definitions of the sense(s) in the
semantic representation
utilized in the particular system, and even the semantic field in
which each sense of a
polysemous word is used.
Syntactic
This level focuses on analyzing the words in a sentence so as to
uncover the grammatical
structure of the sentence. This requires both a grammar and a
parser. The output of this
level of processing is a (possibly delinearized) representation of
the sentence that reveals
the structural dependency relationships between the words.
There are various grammars

that can be utilized, and which will, in turn, impact the choice
of a parser. Not all NLP
applications require a full parse of sentences, therefore the
remaining challenges in
parsing of prepositional phrase attachment and conjunction
scoping no longer stymie
those applications for which phrasal and clausal dependencies
are sufficient. Syntax
conveys meaning in most languages because order and
dependency contribute to
meaning. For example the two sentences: ‘The dog chased the
cat.’ and ‘The cat chased
the dog.’ differ only in terms of syntax, yet convey quite
different meanings.
Semantic
This is the level at which most people think meaning is
determined, however, as we can
see in the above defining of the levels, it is all the levels that
contribute to meaning.
Semantic processing determines the possible meanings of a
sentence by focusing on the
interactions among word-level meanings in the sentence. This

level of processing can
include the semantic disambiguation of words with multiple
senses; in an analogous way
to how syntactic disambiguation of words that can function as
multiple parts-of-speech is
accomplished at the syntactic level. Semantic disambiguation
permits one and only one
sense of polysemous words to be selected and included in the
semantic representation of
the sentence. For example, amongst other meanings, ‘file’ as a
noun can mean either a
folder for storing papers, or a tool to shape one’s fingernails, or
a line of individuals in a
queue. If information from the rest of the sentence were
required for the disambiguation,
the semantic, not the lexical level, would do the
disambiguation. A wide range of
methods can be implemented to accomplish the disambiguation,
some which require
information as to the frequency with which each sense occurs in
a particular corpus of
interest, or in general usage, some which require consideration
of the local context, and

others which utilize pragmatic knowledge of the domain of the
document.
Discourse
While syntax and semantics work with sentence-length units,
the discourse level of NLP
works with units of text longer than a sentence. That is, it does
not interpret multi-
sentence texts as just concatenated sentences, each of which can
be interpreted singly.
Rather, discourse focuses on the properties of the text as a
whole that convey meaning by
making connections between component sentences. Several
types of discourse processing
can occur at this level, two of the most common being anaphora
resolution and
discourse/text structure recognition. Anaphora resolution is the
replacing of words such
as pronouns, which are semantically vacant, with the
appropriate entity to which they
refer (30). Discourse/text structure recognition determines the
functions of sentences in
the text, which, in turn, adds to the meaningful representation

of the text. For example,
newspaper articles can be deconstructed into discourse
components such as: Lead, Main
Story, Previous Events, Evaluation, Attributed Quotes, and
Expectation (31).
Pragmatic
This level is concerned with the purposeful use of language in
situations and utilizes
context over and above the contents of the text for
understanding The goal is to explain
how extra meaning is read into texts without actually being
encoded in them. This
requires much world knowledge, including the understanding of
intentions, plans, and
goals. Some NLP applications may utilize knowledge bases and
inferencing modules. For
example, the following two sentences require resolution of the
anaphoric term ‘they’, but
this resolution requires pragmatic or world knowledge.
The city councilors refused the demonstrators a permit because

they feared
violence.
The city councilors refused the demonstrators a permit because
they advocated
revolution.
Summary of Levels
Current NLP systems tend to implement modules to accomplish
mainly the lower levels
of processing. This is for several reasons. First, the application
may not require
interpretation at the higher levels. Secondly, the lower levels
have been more thoroughly
researched and implemented. Thirdly, the lower levels deal with
smaller units of analysis,
e.g. morphemes, words, and sentences, which are rule-governed,
versus the higher levels
of language processing which deal with texts and world
knowledge, and which are only

regularity-governed. As will be seen in the following section on
Approaches, the
statistical approaches have, to date, been validated on the lower
levels of analysis, while
the symbolic approaches have dealt with all levels, although
there are still few working
systems which incorporate the higher levels.
APPROACHES TO NATURAL LANGUAGE PROCESSING
Natural language processing approaches fall roughly into four
categories: symbolic,
statistical, connectionist, and hybrid. Symbolic and statistical
approaches have coexisted
since the early days of this field. Connectionist NLP work first
appeared in the 1960’s.
For a long time, symbolic approaches dominated the field. In
the 1980’s, statistical
approaches regained popularity as a result of the availability of
critical computational
resources and the need to deal with broad, real-world contexts.
Connectionist approaches
also recovered from earlier criticism by demonstrating the
utility of neural networks in

NLP. This section examines each of these approaches in terms
of their foundations,
typical techniques, differences in processing and system
aspects, and their robustness,
flexibility, and suitability for various tasks.
Symbolic Approach
Symbolic approaches perform deep analysis of linguistic
phenomena and are based on
explicit representation of facts about language through well-
understood knowledge
representation schemes and associated algorithms (21). In fact,
the description of the
levels of language analysis in the preceding section is given
from a symbolic perspective.
The primary source of evidence in symbolic systems comes
from human-developed rules
and lexicons.
A good example of symbolic approaches is seen in logic or rule-
based systems. In logic-
based systems, the symbolic structure is usually in the form of
logic propositions.
Manipulations of such structures are defined by inference

procedures that are generally
truth preserving. Rule-based systems usually consist of a set of
rules, an inference engine,
and a workspace or working memory. Knowledge is represented
as facts or rules in the
rule-base. The inference engine repeatedly selects a rule whose
condition is satisfied and
executes the rule.
Another example of symbolic approaches is semantic networks.
First proposed by
Quillian (10) to model associative memory in psychology,
semantic networks represent
knowledge through a set of nodes that represent objects or
concepts and the labeled links
that represent relations between nodes. The pattern of
connectivity reflects semantic
organization, that is; highly associated concepts are directly
linked whereas moderately or
weakly related concepts are linked through intervening
concepts. Semantic networks are
widely used to represent structured knowledge and have the
most connectionist flavor of
the symbolic models (32).

Symbolic approaches have been used for a few decades in a
variety of research areas and
applications such as information extraction, text categorization,
ambiguity resolution, and
lexical acquisition. Typical techniques include: explanation-
based learning, rule-based
learning, inductive logic programming, decision trees,
conceptual clustering, and K
nearest neighbor algorithms (6; 33).
Statistical Approach
Statistical approaches employ various mathematical techniques
and often use large text
corpora to develop approximate generalized models of linguistic
phenomena based on
actual examples of these phenomena provided by the text
corpora without adding
significant linguistic or world knowledge. In contrast to
symbolic approaches, statistical
approaches use observable data as the primary source of

evidence.
A frequently used statistical model is the Hidden Markov Model
(HMM) inherited from
the speech community. HMM is a finite state automaton that has
a set of states with
probabilities attached to transitions between states (34).
Although outputs are visible,
states themselves are not directly observable, thus “hidden”
from external observations.
Each state produces one of the observable outputs with a certain
probability.
Statistical approaches have typically been used in tasks such as
speech recognition,
lexical acquisition, parsing, part-of-speech tagging,
collocations, statistical machine
translation, statistical grammar learning, and so on.
Connectionist Approach
Similar to the statistical approaches, connectionist approaches
also develop generalized
models from examples of linguistic phenomena. What separates
connectionism from

other statistical methods is that connectionist models combine
statistical learning with
various theories of representation - thus the connectionist
representations allow
transformation, inference, and manipulation of logic formulae
(33). In addition, in
connectionist systems, linguistic models are harder to observe
due to the fact that
connectionist architectures are less constrained than statistical
ones (35); (21).
Generally speaking, a connectionist model is a network of
interconnected simple
processing units with knowledge stored in the weights of the
connections between units
(32). Local interactions among units can result in dynamic
global behavior, which, in
turn, leads to computation.
Some connectionist models are called localist models, assuming
that each unit represents
a particular concept. For example, one unit might represent the
concept “mammal” while
another unit might represent the concept “whale”. Relations

between concepts are
encoded by the weights of connections between those concepts.
Knowledge in such
models is spread across the network, and the connectivity
between units reflects their
structural relationship. Localist models are quite similar to
semantic networks, but the
links between units are not usually labeled as they are in
semantic nets. They perform
well at tasks such as word-sense disambiguation, language
generation, and limited
inference (36).
Other connectionist models are called distributed models.
Unlike that in localist models, a
concept in distributed models is represented as a function of
simultaneous activation of
multiple units. An individual unit only participates in a concept
representation. These
models are well suited for natural language processing tasks
such as syntactic parsing,
limited domain translation tasks, and associative retrieval.

Comparison Among Approaches
From the above section, we have seen that similarities and
differences exist between
approaches in terms of their assumptions, philosophical
foundations, and source of
evidence. In addition to that, the similarities and differences
can also be reflected in the
processes each approach follows, as well as in system aspects,
robustness, flexibility, and
suitable tasks.
Process: Research using these different approaches follows a
general set of steps,
namely, data collection, data analysis/model building, rule/data
construction, and
application of rules/data in system. The data collection stage is
critical to all three
approaches although statistical and connectionist approaches
typically require much more
data than symbolic approaches. In the data analysis/model
building stage, symbolic
approaches rely on human analysis of the data in order to form a
theory while statistical

approaches manually define a statistical model that is an
approximate generalization of
the collected data. Connectionist approaches build a
connectionist model from the data.
In the rule / data construction stage, manual efforts are typical
for symbolic approaches
and the theory formed in the previous step may evolve when
new cases are encountered.
In contrast, statistical and connectionist approaches use the
statistical or connectionist
model as guidance and build rules or data items automatically,
usually in relatively large
quantity. After building rules or data items, all approaches then
automatically apply them
to specific tasks in the system. For instance, connectionist
approaches may apply the
rules to train the weights of links between units.
System aspects: By system aspects, we mean source of data,
theory or model formed
from data analysis, rules, and basis for evaluation.
- Data: As mentioned earlier, symbolic approaches use human
introspective data, which

are usually not directly observable. Statistical and connectionist
approaches are built on
the basis of machine observable facets of data, usually from text
corpora.
- Theory or model based on data analysis: As the outcome of
data analysis, a theory is
formed for symbolic approaches whereas a parametric model is
formed for statistical
approaches and a connectionist model is formed for
connectionist approaches.
- Rules: For symbolic approaches, the rule construction stage
usually results in rules with
detailed criteria of rule application. For statistical approaches,
the criteria of rule
application are usually at the surface level or under-specified.
For connectionist
approaches, individual rules typically cannot be recognized.
- Basis for Evaluation: Evaluation of symbolic systems is
typically based on intuitive
judgments of unaffiliated subjects and may use system-internal

measures of growth such
as the number of new rules. In contrast, the basis for evaluation
of statistical and
connectionist systems are usually in the form of scores
computed from some evaluation
function. However, if all approaches are utilized for the same
task, then the results of the
task can be evaluated both quantitatively and qualitatively and
compared.
Robustness: Symbolic systems may be fragile when presented
with unusual, or noisy
input. To deal with anomalies, they can anticipate them by
making the grammar more
general to accommodate them. Compared to symbolic systems,
statistical systems may be
more robust in the face of unexpected input provided that
training data is sufficient,
which may be difficult to be assured of. Connectionist systems
may also be robust and
fault tolerant because knowledge in such systems is stored
across the network. When
presented with noisy input, they degrade gradually.

Flexibility: Since symbolic models are built by human analysis
of well-formulated
examples, symbolic systems may lack the flexibility to adapt
dynamically to experience.
In contrast, statistical systems allow broad coverage, and may
be better able to deal with
unrestricted text (21) for more effective handling of the task at
hand. Connectionist
systems exhibit flexibility by dynamically acquiring appropriate
behavior based on the
given input. For example, the weights of a connectionist
network can be adapted in real-
time to improve performance. However, such systems may have
difficulty with the
representation of structures needed to handle complex
conceptual relationships, thus
limiting their abilities to handle high-level NLP (36).
Suitable tasks: Symbolic approaches seem to be suited for
phenomena that exhibit
identifiable linguistic behavior. They can be used to model
phenomena at all the various
linguistic levels described in earlier sections. Statistical
approaches have proven to be

effective in modeling language phenomena based on frequent
use of language as reflected
in text corpora. Linguistic phenomena that are not well
understood or do not exhibit clear
regularity are candidates for statistical approaches. Similar to
statistical approaches,
connectionist approaches can also deal with linguistic
phenomena that are not well
understood. They are useful for low-level NLP tasks that are
usually subtasks in a larger
problem.
To summarize, symbolic, statistical, and connectionist
approaches have exhibited
different characteristics, thus some problems may be better
tackled with one approach
while other problems by another. In some cases, for some
specific tasks, one approach
may prove adequate, while in other cases, the tasks can get so
complex that it might not
be possible to choose a single best approach. In addition, as
Klavans and Resnik (6)
pointed out, there is no such thing as a “purely statistical”
method. Every use of statistics

is based upon a symbolic model and statistics alone is not
adequate for NLP. Toward this
end, statistical approaches are not at odds with symbolic
approaches. In fact, they are
rather complementary. As a result, researchers have begun
developing hybrid techniques
that utilize the strengths of each approach in an attempt to
address NLP problems more
effectively and in a more flexible manner.
NATURAL LANGUAGE PROCESSING APPLICATIONS
Natural language processing provides both theory and
implementations for a range of
applications. In fact, any application that utilizes text is a
candidate for NLP. The most
frequent applications utilizing NLP include the following:
• Information Retrieval – given the significant presence of text
in this application, it is
surprising that so few implementations utilize NLP. Recently,
statistical approaches

for accomplishing NLP have seen more utilization, but few
systems other than those
by Liddy (37) and Strzalkowski (38) have developed significant
systems based on
NLP
.
• Information Extraction (IE) – a more recent application area,
IE focuses on the
recognition, tagging, and extraction into a structured
representation, certain key
elements of information, e.g. persons, companies, locations,
organizations, from large
collections of text. These extractions can then be utilized for a
range of applications
including question-answering, visualization, and data mining.
• Question-Answering – in contrast to Information Retrieval,
which provides a list of
potentially relevant documents in response to a user’s query,
question-answering
provides the user with either just the text of the answer itself or
answer-providing
passages.

• Summarization – the higher levels of NLP, particularly the
discourse level, can
empower an implementation that reduces a larger text into a
shorter, yet richly-
constituted abbreviated narrative representation of the original
document.
• Machine Translation – perhaps the oldest of all NLP
applications, various levels of
NLP have been utilized in MT systems, ranging from the ‘word-
based’ approach to
applications that include higher levels of analysis.
• Dialogue Systems – perhaps the omnipresent application of the
future, in the systems
envisioned by large providers of end-user applications.
Dialogue systems, which
usually focus on a narrowly defined application (e.g. your
refrigerator or home sound
system), currently utilize the phonetic and lexical levels of
language. It is believed
that utilization of all the levels of language processing
explained above offer the
potential for truly habitable dialogue systems.

CONCLUSIONS
While NLP is a relatively recent area of research and
application, as compared to other
information technology approaches, there have been sufficient
successes to date that
suggest that NLP-based information access technologies will
continue to be a major area
of research and development in information systems now and
far into the future.
Acknowledgement
Grateful appreciation to Xiaoyong Liu who contributed to this
entry while she was a
Ph.D. student and a Research Assistant in the Center for Natural
Language Processing
in the School of Information Studies at Syracuse University.

Syracuse UniversitySURFACE2001Natural Language
ProcessingElizabeth D. LiddyRecommended CitationMicrosoft
Word - 03NLP.LIS.Encyclopedia.doc.doc
Professional Psychology: Research and
Practice
Artificial Intelligence in Psychological Practice: Current
and Future Applications and Implications
David D. Luxton
Online First Publication, November 11, 2013. doi:
10.1037/a0034559
CITATION
Luxton, D. D. (2013, November 11). Artificial Intelligence in
Psychological Practice: Current
and Future Applications and Implications. Professional
Psychology: Research and Practice.
Advance online publication. doi: 10.1037/a0034559
Artificial Intelligence in Psychological Practice: Current and
Future
Applications and Implications
David D. Luxton
National Center for Telehealth & Technology, Tacoma,
Washington
and University of Washington School of Medicine, Seattle
This article reviews developments in artificial intelligence (AI)

technologies and their current and
prospective applications in clinical psychological practice.
Some of the principal AI assisted activities
reviewed include clinical training, treatment, psychological
assessment, and clinical decision making. A
concept for an integrated AI-based clinician system is also
introduced. Issues associated with AI in the
context of clinical practice, the potential for job loss among
mental health professionals, and other
ramifications associated with the advancement of AI technology
are discussed. The advancement of AI
technologies and their application in psychological practice
have important implications that can be
expected to transform the mental health care field.
Psychologists and other mental health care profes-
sionals have an essential part to play in the development,
evaluation, and ethical use of AI technologies.
Keywords: artificial intelligence, mental health, expert systems,
virtual reality
Artificial Intelligence (AI) is technology designed to perform
activities that normally require human intelligence. AI is also
defined as the multidisciplinary field of science that is
concerned
with the development and study of this technology. The field of
AI
finds its genesis with the beginning of the computer age in the
1940s, and it was officially given its name by computer scientist
John McCarthy in 1956 (see Buchanan, 2005, for a review of
the
history of AI). AI technology can be in the form of physical
machines, standalone computer software, distributed across net-
works, applied to robotics, or engineered from living biology or
coupled to it (e.g., brain– computer interfaces). This technology
can be purposed for specialized intelligent functions or to

emulate
complex human behavior that is capable of reasoning, learning,
and acting upon an environment as an autonomous intelligent
agent (Russell & Norvig, 2003). Important branches of AI
include
the study of machine learning, artificial neural networks, and
natural language processing. Machine learning is the ability of
computers to learn without being explicitly programmed
(Samuel,
1959), artificial neural networks are mathematical,
computational,
or technological models that mimic the logic and learning func-
tions of neurons in a brain (Krogh, 2008), and natural language
processing is concerned with how computers process human nat-
ural languages (Manning & Schütze, 1999).
AI has been applied to activities in the field of medicine since
the 1970s, particularly in the areas of expert systems for
clinical
decision making and in biomedical research (Morelli, Bronzino
& Goethe, 1987; Patel et al., 2009; Shortliffe, 1993; Szolovits,
1982). The emergence of AI in medicine has also brought forth
the scientific journal “Artificial Intelligence in Medicine” and
several earlier reviews and proposals of AI applications in
psychiatry have been published (e.g., Garfield, Rapp, & Evens,
1992; Hand, 1985; Morelli, 1989; Servan-Schreiber, 1986). The
use of AI technologies in the mental health care field remains a
burgeoning area that has seen important developments in the
last decade. The steady increase in computer performance as
well as advances in other technological areas such as in virtual
reality, computer knowledge acquisition, language processing,
sensing, and robotics have enabled new and exciting capabili-
ties that were only dreamed of in the past. The current and
forthcoming applications of AI technologies can be expected to
have a profound impact on the field of psychology and mental

health care in general. It is therefore important for psycholo-
gists and others in the mental health care field to be aware of
the
both the capabilities and ramifications of the use of current and
emerging AI technologies.
The focus of this article is therefore to review the uses of AI
technologies that are applicable to activities in psychological
prac-
tice and research. It is not feasible to present an exhaustive
review
of all AI technologies or applications in this article, however
illustrative examples of AI technology applications that are cur-
rently being used or evaluated are described. Basic historical
background and technical descriptions are provided for readers
who are new to the topic and prospects and possibilities for
future
AI technology applications are presented. Finally, the
implications
DAVID D. LUXTON is a licensed clinical psychologist who
received his PhD
in clinical psychology from the University of Kansas. He is a
Research
Psychologist and Program Manager at National Center for
Telehealth &
Technology (T2) and an Affiliate Associate Professor of
Psychiatry and
Behavioral Sciences at the University of the Washington School
of Med-
icine in Seattle. His research and writing are focused in the
areas of military
psychological health, telehealth, and technology-based
treatments.
THE VIEWS EXPRESSED are those of the author and do not
reflect the official

policy or position of the Department of Defense of the U.S.
Government.
CORRESPONDENCE CONCERNING THIS ARTICLE should
be addressed to David
D. Luxton, National Center for Telehealth & Technology (T2),
Defense
Centers of Excellence (DCoE) for Psychological Health &
Traumatic Brain
Injury, 9933 West Hayes Street, Joint Base Lewis-McChord,
WA 98431.
E-mail: [email protected] or [email protected]
T
hi
s
do
cu
m
en
t
is
co
py
ri
gh
te
d
by
th

e
A
m
er
ic
an
P
sy
ch
ol
og
ic
al
A
ss
oc
ia
ti
on
or
on
e
of
it

s
al
li
ed
pu
bl
is
he
rs
.
T
hi
s
ar
ti
cl
e
is
in
te
nd
ed
so
le

ly
fo
r
th
e
pe
rs
on
al
us
e
of
th
e
in
di
vi
du
al
us
er
an
d
is

no
t
to
be
di
ss
em
in
at
ed
br
oa
dl
y.
Professional Psychology: Research and Practice In the public
domain
2013, Vol. 44, No. 6, 000 DOI: 10.1037/a0034559
1
of the advancement of this technology for patients, mental
health
professionals, and the field of psychology are discussed.
Clinical Treatment, Assessment, and Training

The first simulation of a psychotherapist that used a human-
computer interface was the ELIZA computer program in 1966
(Weizenbaum, 1976). The program was designed to imitate the
empathic communication style of Carl Rogers (Rogers, 1951),
and
it used a question and answer format to respond to statements
that
its user typed on a keyboard. ELIZA used language syntax to
provide formulated responses based a programmed model and
therefore only mimicked conversation. In the early 1970s,
psychi-
atrist Kenneth M. Colby developed a program called PARRY at
Stanford University that simulated a person with paranoid
schizo-
phrenia and, like ELIZA, the program could converse with
others
(Güzeldere & Franchi, 1995). PARRY is credited as being the
first
program to pass the Turing Test. The Turing Test (Turing,
1950),
named after Alan Turing, is a method for judging the
intelligence
of machines. To pass the test, a computer program must imper-
sonate a human real-time written conversation with a human
judge
sufficiently enough so that the judge cannot reliably distinguish
between the program and a real person. Tests of PARRY showed
that expert psychiatrists were unable to distinguish between
PARRY and an actual person with paranoid schizophrenia
(Teuscher & Hofstadter, 2006).
Technology has now developed into advanced virtual human
avatars (virtual reality simulated humans) that are capable of
carrying on interactive and intelligent conversations. The
coupling
of virtual reality simulation, natural language processing, and

knowledge-based AI capable of reasoning makes this possible.
Researchers at University of Southern California’s (USC)
Institute
for Creative Technologies, for example, are currently
developing
life-like virtual human patients for use in clinical training and
skill
acquisition (Rizzo, Lange, et al., 2011). The virtual reality
patients
are designed to mimic the symptoms of psychological disorders
and interact with therapists through verbal dialogue. They can
also
be modified for specific patient population simulations and
trainee
skill levels. Some of the potential benefits of this technology
include the capability for trainees to receive adaptive and
custom-
ized training that is highly realistic and also available to the
trainee
at any time. This can provide the added benefit of freeing up
instructors to play a more advanced role in guiding student
train-
ing. More research is needed, however, to determine how
effective
these systems will be.
AI-enabled virtual reality human avatars have the potential to
be
used for all other types of person-to-person interactions in
mental
health care including psychological treatments, assessments,
and
testing. The use of virtual reality avatars to provide people with
information about mental health resources and support are
already
in use (DeAngelis, 2012; Rizzo, Lange, et al., 2011). SimCoach

(www.simcoach.org), for example, is designed to connect
military
service members and their families to health care and other
well-
being resources (Rizzo, Lange, et al., 2011). This type of AI
technology may one day revolutionize telepractice—AI-enabled
avatars could be accessed remotely to provide psychological
ser-
vices to anywhere where there is an Internet connection. One of
the
benefits for patients is that these automated AI-enabled virtual
consultants can be conveniently accessed by patients at any time
and provide them with basic assessments, recommendations, and
referrals for further treatment that are tailored to the patient’s
individual needs. Another advantage of virtual reality avatar
sys-
tems is that persons who are concerned about privacy and the
stigma associated with seeking care in person may be more
willing
to seek help from a virtual care provider in the comfort of their
home. Another benefit of this technology is that it is more inter-
active and engaging than static informational Internet Web
sites.
These systems also have the potential to assist practitioners by
serving as always available specialist consultants that have
learned
and possess knowledge in particular domains or disciplines.
The use of AI-enabled kiosk-based computerized health screen-
ing systems may also be advantageous in settings where large
numbers of people need to be screened, such as in the military.
Systems that use AI machine learning and reasoning concepts
go
beyond mere computerized surveys with logic-based algorithms
and gate questions; they could make assessments more efficient

and sophisticated because of the capability to process complex
data, customize to the individual, and reduce uncertainty in
screen-
ing outcomes.
The Super Clinician
Integrated AI technologies can also provide a simulated practi-
tioner with capabilities that are beyond those of human
practitio-
ners, effectively making it a super clinician. The super clinician
could be built with advanced sensory technologies such as
infrared
imaging (to detect body temperature changes indicative of
changes
in internal states) and optical sensing capable of observing and
analyzing subtle facial expressions, eye blinking, vocal
character-
istics, and other patterns of behavior that provide clinically
rele-
vant information. Machine olfaction technology could also be
used
to sense the presence of alcohol, for example. The technology
could use facial recognition technology to verify the identity of
patients and also access and analyze all data available about the
patient from electronic medical records, session notes, assess-
ments, and testing results via wireless technologies.
Furthermore,
the super clinician could conduct sessions with complete auton-
omy or serve as an assistant to practitioners during clinical
assess-
ments and treatments. For example, this technology could assist
the human practitioner with records review, monitoring of
physi-
ological data, pretreatment clinical interviews, or test
administra-

tion.
As evidenced by several projects in this area, the super clinician
concept is not science fiction fantasy. For example, USC’s
Insti-
tute for Creative Technologies’ work on the Defense Advanced
Research Projects Agency (DARPA) Detection and
Computational
Analysis of Psychological Signals (DCAPS) project involves
de-
velopment of an AI system that uses machine learning, natural
language processing, and computer vision to analyze language,
physical gestures, and social signals to detect psychological dis-
tress cues in humans (DARPA, 2013). Researchers at the Massa-
chusetts Institute of Technology (MIT) Computer Science and
Artificial Intelligence Laboratory (CSAIL) have designed
software
that amplifies variations in digital video pixels that allows the
observation of subtle changes that are not noticeable to the
human
eye (Hardesty, 2012). This technology could be used to detect a
person’s pulse rate (i.e., internal arousal states) as the skin
color
changes with the flow of blood. Also, Watson is IBM’s AI lan-
T
hi
s
do
cu
m
en
t

is
co
py
ri
gh
te
d
by
th
e
A
m
er
ic
an
P
sy
ch
ol
og
ic
al
A
ss

oc
ia
ti
on
or
on
e
of
it
s
al
li
ed
pu
bl
is
he
rs
.
T
hi
s
ar
ti

cl
e
is
in
te
nd
ed
so
le
ly
fo
r
th
e
pe
rs
on
al
us
e
of
th
e
in

di
vi
du
al
us
er
an
d
is
no
t
to
be
di
ss
em
in
at
ed
br
oa
dl
y.

2 LUXTON
guage processing question answering system that defeated Jeop-
ardy! quiz show champions Brad Rutter and Ken Jennings
during
an exhibition match in 2011. IBM is currently evaluating an
expanded, commercially available version of Watson that has
learned the medical literature, therefore allowing it to serve as a
medical knowledge expert and consultant (IBM, 2013). The
FDA
recently approved a robot called RP-VITA for use in hospitals
that
can maneuver from room to room to connect health care
providers
to patients or to other health care providers via wireless video
teleconferencing (InTouch Health, 2012). The system can also
access the medical records of patients and can be used to
monitor
patients remotely. Undoubtedly, the continued advancement,
inte-
gration, and application of these types of technologies will
create
opportunities to build intelligent agent systems that are capable
of
providing the range of psychological treatment, assessment, and
education services.
Clinical Diagnostics and Decision Making
One of the earliest applications of computer and AI technology
in the medical field that also has direct applicability to the
mental
health care field is the use of expert systems for clinical
decision

making. An expert system is a computer program designed to
incorporate the knowledge and ability of an expert in a
particular
domain (McCarthy, 1984), and decision support systems are a
class of expert system that is specifically designed to aid in the
process of decision making (Finlay, 1994). Many of these
systems
are rule-based expert systems that have facts and rules prepro-
grammed and therefore require a priori knowledge on the part of
the decision maker. Decision support systems can also be
designed
to use data mining techniques to search and find patterns and
relationships in data and therefore do not require a priori
knowl-
edge (Hardin & Chhien, 2007). Also, fuzzy expert systems are
expert systems that use fuzzy logic instead of Boolean logic.
Fuzzy
logic (Zadeh, 1965) is a method of reasoning that deals with
approximate values (e.g., some degree of “true”) rather than
fixed
and exact values (e.g., “true” or “false”) and is useful for
working
with uncertainties during decision making. Fuzzy modeling and
fuzzy-genetic algorithms are techniques used to assist with the
optimization of rules and membership classification (see
Jagielska,
Matthews & Whitfort, 1999 for a review of these concepts).
One of the first clinical decision support programs was devel-
oped at Stanford University in the early 1970s. The system,
known
as MYCIN, was designed to identify bacteria causing infections
and blood clotting diseases (Buchanan & Shortliffe, 1984;
Short-
liffe, 1976). Built by interviewing experts, MYCIN was a rule-
based system that used a typed question and answer dialog. Al-

though the system performed well in tests, it was never put to
clinical use mostly because of the computing technology limita-
tions of the day (Buchanan & Shortliffe, 1984). The
advancements
in computing power and AI technology since then, however,
have
greatly improved the capabilities of clinical expert systems.
With
the use of neural network concepts and machine learning tech-
niques, modern expert systems can identify patterns, trends, and
meaning from complex data that are too complex to be
processed
by humans or other computer-based technologies. Support
vector
machines (SVMs; Cortes & Vapnik, 1995), for example, use
machine learning to analyze, classify, and recognize patterns in
data and have recently been tested in the classification of
several
diseases including Parkinson’s disease (Gil & Johnson, 2009)
and
Alzheimer’s disease (Kohannim et al., 2010).
The use of expert systems in the mental health field has lagged
behind application in the medical field, however the
applicability
of AI enhanced systems is being realized. For example, Masri
and
Mat Jani (2012) proposed an AI-based Mental Health Diagnostic
Expert System (MeHDES) that would encode human experts’
knowledge of mental health disorders into a knowledge base
using
rule-based reasoning techniques. Fuzzy logic techniques would
then be used to determine the severity level of a particular
disorder
to be measured, and fuzzy-genetic algorithms would be used to

determine and propose personalized treatments that consider the
patient’s budget and overall health condition. AI-enabled virtual
reality human avatars with speech detection and natural
language
processing technology could also enhance expert systems by
pro-
viding a human-like verbal dialogue interface. These systems
could have access to the corpus of expert knowledge regarding
psychiatric and medical disorders and be fed data from patient
medical records and testing results. Other practical applications
of
AI-enabled expert systems include assistance with review of
med-
ications use, monitoring, and identification of contraindications
(Bindoff, Stafford, Peterson, Kang, & Tenni, 2012). Moreover,
the
concept of artificial intelligent multiagents could also be used
to
have artificial “minds” work collectively to make decisions and
provide solutions to problems in clinical practice or research.
Along these lines, McShane, Beale, Nirenburg, Jarell, and
Fantry
(2012) discuss a system that enables the creation of artificial
intelligent agents that can operate as members of multiagent
teams
(i.e., both artificial and human medical experts) to detect and
resolve medical diagnostic inconsistencies.
The benefit of AI-based clinical decision support systems is that
they can deal with high levels of complexity in data and can
therefore assist practitioners with extracting relevant
information
and making optimal decisions. These systems can also help
prac-
titioners deal with uncertainty and help speed up decision
making.

The application of AI-enabled clinical decision support systems
can reduce demands on staff time and it can help reduce barriers
of
limited practitioner competence in particular areas. Moreover,
as
humans are susceptible to making mistakes as a result of
cognitive
errors and fatigue, AI technology has the potential to enhance
capabilities and reduce human errors in clinical decision making
in
all health care fields.
Other Practical AI Applications
in Mental Health Care
Intelligent Virtual Worlds and Artificial Companions
Virtual reality simulation is also an emerging application of AI
technologies. Virtual reality is a form of human– computer
inter-
face that allows the user to become immersed within and
interact
with a computer-generated simulated environment (Rizzo, Buck-
walter & Neumann, 1997). Clinical virtual reality is the use of
this
technology for clinical assessment and treatment purposes
(Rizzo,
Parsons, et al., 2011), and it has been used in the treatment of a
variety of psychological disorders (see Gorrindo, & Groves,
2009;
Krijn, Emmelkamp, Olafsson, & Biemond, 2004; Reger, Hollo-
way, Rothbaum, Difede, & Gahm, 2011; Riva, 2010). AI is
already
T
hi

s
do
cu
m
en
t
is
co
py
ri
gh
te
d
by
th
e
A
m
er
ic
an
P
sy
ch

ol
og
ic
al
A
ss
oc
ia
ti
on
or
on
e
of
it
s
al
li
ed
pu
bl
is
he
rs

.
T
hi
s
ar
ti
cl
e
is
in
te
nd
ed
so
le
ly
fo
r
th
e
pe
rs
on
al

us
e
of
th
e
in
di
vi
du
al
us
er
an
d
is
no
t
to
be
di
ss
em
in
at

ed
br
oa
dl
y.
3ARTIFICIAL INTELLIGENCE IN PSYCHOLOGICAL
PRACTICE
used in virtual environments to create intelligent agents that can
learn and interact with users and therefore increase flexibility
and
realism. Further, these artificial intelligent agents are now able
to
express emotion and participate in dialogue with human users.
Also, “biologically inspired” virtual companions, such as virtual
household pets, may have mental health benefits by promoting
mental wellbeing and helping people to cope with loneliness.
These can be in virtual form, appearing on a video screen, or in
the
form of animal or humanoid robots. For example, animal robot
companions have been designed to provide therapy for patients
with dementia (see Shibata & Wada, 2011). Just as with AI-
enhanced video games, AI makes these artificial companions
more
life-like, interactive, and capable of doing things that are
adaptive
to a patient’s needs.
Augmented Reality Applications

Augmented reality combines virtual reality with the real world
by superimposing computer generated graphics with live video
imagery (Caudell & Mizell, 1992). This technology, when com-
bined with other AI technologies, could transform how humans
perceive and interact within their environments and could be
used
for a variety of therapeutic purposes. For example, it could be
used
to create anxiety provoking virtual stimuli in the patient’s real-
world environment during prolonged exposure therapy or be
used
to assist patients with real-time therapeutic virtual coaching that
is
projected on the screen. Augmented reality and other AI
capabil-
ities can also be applied to mobile devices such as smartphones,
tablet PCs, and other wearable devices. For example, Google’s
Glass (wearable intelligent glasses) can provide users with
access
to the Internet for real-time data access and sharing and other
capabilities. Researchers at the University of Washington and
Aalto University (Finland) are also currently developing bionic
contact lenses that may one day lead to technology that enables
users to scan the Internet and have access to data on demand,
such
as medical information (Lingley et al., 2011).
Therapeutic Computer Games
Computer games can be used for mental health care purposes
such as skills training, behavior modeling, therapeutic
distraction,
and other therapeutic purposes. Some of the therapeutic benefits
of
computer games include increased engagement of patients, im-
proved adherence to treatments, and reduced stigma associated

with psychological treatment (Matthews & Coyle, 2010). Thera-
peutic computer games have also been shown to help
adolescents
improve self-confidence and problem solving skills (Coyle,
Mathews, Sharry, Nisbet, & Doherty, 2005). AI technology is
already present in many commercial computer games and has
more recently been applied to Internet-based online and social
network games (Fujita & Wu, 2012). AI and machine learning
technology, when applied to computer games, enhances realism,
which makes the games more interesting, challenging, and
enter-
taining for game play. Machine learning concepts also help
make
the games customizable to the patient’s needs. That is, AI tech-
nology can be used to direct game play so that the patient
practices
skills in needed areas or patients can be coached by virtual
intel-
ligent agents within games or other virtual environments such as
Second Life (Linden Research, Inc., 2013). Brigadoon (Lester,
2005), for example, is virtual environment in Second Life that
is
designed for people with autism spectrum disorder. The
simulation
allows users to interact with avatars to learn and practice social
skills in a nonthreatening environment.
Other Clinical Tools
The integration of AI into other clinical tools that mental health
care and other medical professionals use can increase
convenience,
accuracy, and efficiency. The use of speech recognition technol-
ogy for medical dictation has been used for some time. There
now

exist, however, electronic medical record (EMR) software appli-
cations that use AI and Boolean logic to automate patient data
entry by recalling elements from past cases that are the same or
similar to the case thereby improving accuracy and saving time.
Another
application may be an AI-based program that listens to the
therapy or
assessment session and intelligently summarizes the session
auto-
matically, essentially eliminating the need to make clinical
chart
notes at session end. This type of system could be implemented
on
mobile device platforms such as smartphones.
Implications of AI in Psychological Practice
Interaction Between Humans and AI
The use of AI systems in the context of clinical interaction with
patients raises a number of very intriguing questions. For
example,
will patients be able to develop therapeutic bonds and trust with
artificial intelligent agent practitioners as they may with human
practitioners? How will patients interact with artificial
intelligent
agent practitioners if the patient thinks that the system lacks the
imperfections of humans or is using advanced technologies that
exceed the knowledge and sensory capabilities of humans?
Joseph
Weizenbaum, the creator of ELIZA program, argued that
comput-
ers should not be allowed to make important decisions because
computers lack the human qualities of compassion and wisdom
(Weizenbaum, 1976). Others have argued, however, that AI-
enabled machines can indeed experience emotions, or at least

the
recognition and expression of emotions can be modeled in a
machine (Bartneck, Lyons, & Saerbeck, 2008). Interpersonal
warmth, empathy, and the therapeutic relationship are important
common factors that influence therapeutic outcomes (Lambert &
Barley, 2001). Moreover, cultural differences and expectations
are
also relevant to psychological practice. Even if specific therapy
techniques are appropriately administered by artificial
intelligent
agent practitioners, these common factors and cultural aspects
need to be considered in any discussions about how these
systems
should be used in the context of psychotherapy and whether
they
will be effective at treating patients. These questions point to
the
need for research in this area.
Legal and Ethical Considerations
The application of artificial intelligent agent systems to provide
treatment services brings new complexities to the legal and
ethical
issues associated with psychological practice. For example, sys-
tems that are accessible via the Internet, such as current avatar
systems, can provide services across jurisdictional boundaries
(state and national lines). Although these systems are typically
T
hi
s
do
cu

m
en
t
is
co
py
ri
gh
te
d
by
th
e
A
m
er
ic
an
P
sy
ch
ol
og
ic

al
A
ss
oc
ia
ti
on
or
on
e
of
it
s
al
li
ed
pu
bl
is
he
rs
.
T
hi

s
ar
ti
cl
e
is
in
te
nd
ed
so
le
ly
fo
r
th
e
pe
rs
on
al
us
e
of

th
e
in
di
vi
du
al
us
er
an
d
is
no
t
to
be
di
ss
em
in
at
ed
br
oa

dl
y.
4 LUXTON
used for educational purposes with appropriate disclaimers
regard-
ing their use, other treatment applications and contexts may in-
volve the same legal and professional licensure considerations
associated with current telepractice (see Kramer, Mishkind,
Lux-
ton, & Shore, 2013 for a review). The use of advanced
autonomous
AI systems to provide treatment or assessment services,
however,
complicates the liability issues associated with the provision of
services. To deal with ethical dilemmas that health care profes-
sionals face in their everyday practice, artificial intelligent
agents
must be able to process and make value decisions and judgments
that involve complex abstract thinking and reasoning. Although
AI
systems may help improve decision making, just as with a
human
practitioner, AI systems are susceptible to errors of judgment
and
incorrect assessment of risk (e.g., level of self-harm risk for a
patient). Moreover, advanced artificial intelligent agents may be
capable of developing their own personal values and beliefs that
inform decisions—which raises the question of whether those
decisions will be consistent with those of their creators or the
cultural context of use. These types of questions raise concerns
about who should be legally responsible for the decisions and

any
mistakes made by AI systems. Although it seems logical that the
responsibility will ultimately be upon the human controllers of
the
AI system, the question of responsibility certainly becomes
blurred
with the use of autonomous AI systems.
Indeed, the advancement of AI technology has many moral and
ethical considerations associated with the actions of humans
who
control the technology as well as with intelligent machines that
function autonomously (see Anderson & Anderson, 2011).
Science
fiction author Isaac Asimov proposed ethical guidelines
regarding
the use of artificially intelligent machines in the 1940s with his
groundbreaking “Three Laws of Robotics” (Asimov, 1942). In
brief, the laws state that artificially intelligent robots must not
harm a human being, must obey orders of human beings (unless
in
conflict with the first law), and they must protect their own
existence (as long as this does not conflict with the second law).
Asimov later added a preceding law stating that a robot should
not
harm humanity (Asimov, 1985). The need for guidelines
regarding
the ethical use of AI is no longer a matter of science fiction or
philosophy, but a real-world practical issue that is relevant to
professionals in the mental health care field. Further legal dis-
course and guidelines are needed and can be expected in the
future.
Job Loss in Mental Health Care
Although the field of psychology has always adapted to and

made use of the technological innovations of the era, AI innova-
tions are especially significant because they not only improve
and
advance psychological practice and research, but have the
potential
to supplant mental health care professionals in core activities
that
require human intelligence and social interaction. The displace-
ment of workers due to AI enabled systems and other technolog-
ical innovations is already occurring in the banking sector,
semi-
conductor design, customer service jobs, and in the law
profession
to name a few (Brynjolfsson & McAfee, 2011; Markoff, 2011).
The mental health care profession is certainly not immune to
this
risk. Clinical psychologists, for example, will spend upward of
a
decade in college, graduate school, internship, and postdoctoral
experiences to obtain knowledge and learn the skills of the pro-
fession. AI-enabled systems, such as Watson, are capable of
scan-
ning all digitized knowledge and nearly instantaneously
analyzing,
reasoning, and making decisions based on it. This technology
can
certainly be applied to any knowledge-based profession,
including
Clinical Psychology. Moreover, autonomous artificial
intelligent
agents with human-like social capabilities are already able to
interact with people, learn from real-world experiences, and
per-
haps one day conduct the full range of mental health services.
Although it is doubtful psychologists and other mental health

professionals will be replaced by virtual artificial intelligent
agents
or AI-enabled robots any time in the near future, the use of AI
technologies can be expected to have an economic impact on
psychological services in the years ahead.
The Effects of Cognitive Enhancement
The coupling of AI technology directly to the human brain has
already emerged in the medical field as a way to repair and
assist
human cognitive or sensory-motor functions. For example,
direct
brain implants have already been used to control prosthetic
limbs
(Wolpaw, Birbaumer, McFarland, Pfurtscheller, & Vaughan,
2002), treat noncongenital (acquired) blindness (Naam, 2010),
and
in China, tested as a way to help physically challenged people
write Chinese characters (Minett et al., 2012). Brain Computer
Interfaces (BCIs) have also been used for nonmedical purposes
to
communicate with and control devices (Wolpaw, Birbaumer,
Mc-
Farland, Pfurtscheller, Vaughan, 2002).
Implanted AI technologies also have the potential to repair or
improve general cognitive abilities in humans by making people
into cyborgs (partly human and partly machine) (Kurzweil,
2005;
Naam, 2010). This technology may one day provide the benefit
of
restoring function to areas in the brain that have become
damaged
by strokes, traumatic brain injuries, or other organic disorders.
The

technology could also be used to provide patients with real-time
biofeedback and could be used to control the automatic release
of
medical nanotechnologies or psychotropic medications at
prepro-
grammed times or upon specific situational cues such as the
presence of stress or other stimuli. The advancement of this
tech-
nology, however, may have unintended psychological and social
implications. For example, the possession of cognitive enhance-
ments may alter one’s sense of self and behavior in unexpected
ways. Moreover, the belief that others may have particular
cogni-
tive advantages over others may create states of anxiety and
mistrust. The study of the psychological effects of AI enhanced
capabilities on the individual and on groups of people is an area
of
research that psychologists may most certainly contribute.
Artificial Intelligence Superiority
One of the most interesting questions is if and when AI will
have the capability to fully emulate the human brain. The term
Strong AI, introduced by John Searle in 1980 (Searle, 1980), is
a
category of AI that aims to build machines with intellectual
ability
that is indistinguishable from that of human beings. Although
reproducing human general intelligence may still be beyond the
reach of AI at this time, technological advances are closing the
gap
at an incredible pace. Some believe that work in Strong AI will
lead to computers with intelligence that surpasses that of human
beings (Kurzweil, 2005; Vinge, 1993). Ray Kurzweil, futurist
and
Director of Engineering at Google, predicts that this will occur

by
T
hi
s
do
cu
m
en
t
is
co
py
ri
gh
te
d
by
th
e
A
m
er
ic
an

P
sy
ch
ol
og
ic
al
A
ss
oc
ia
ti
on
or
on
e
of
it
s
al
li
ed
pu
bl

is
he
rs
.
T
hi
s
ar
ti
cl
e
is
in
te
nd
ed
so
le
ly
fo
r
th
e
pe

rs
on
al
us
e
of
th
e
in
di
vi
du
al
us
er
an
d
is
no
t
to
be
di
ss

em
in
at
ed
br
oa
dl
y.
PRACTICE
2029 (Kurzweil, 2005). Kurzweil’s prediction is partly based on
Moore’s Law (Moore, 1965), which has reliably demonstrated
that
both the speed and memory capacity of computers double every
two years. He also predicts that by 2045, AI technology will
have
exponentially advanced and improved itself to a point called the
singularity (Kurzweil, 2005; Vinge, 1993; von Neumann, 2012).
Akin to the use of the term in astrophysics to describe the un-
knowns associated with effects of gravity in black holes, the
singularity refers to the unpredictability of what will happen at
that
transformative point in human history when machines develop
superintelligence.
There are indeed unknown outcomes associated with technology
that approaches human general intelligence—and exceeds it.
One

possibility is that the advancement of AI technology will allow
machines to develop their own teleology not conceived by their
creators. Although it is not likely that AI technology will be
allowed to evolve into insidious intelligent agents that aim to
take
over the world, a more immediate concern involves how this
technology will be implemented, controlled, and whether its ap-
plication will be used for the best interest and wellbeing of the
general population. Similar to the development of nuclear
technol-
ogy in the 1940s, humankind is again creating something that
wields great power that once created, there is no turning back.
Nonetheless, advances in AI technology will continue to bring
incredible possibilities and opportunities that have the potential
to
improve the world if approached with wisdom and beneficence.
Conclusion
The presence of AI technology can already be found all around
us; it is used in logistics planning, finance (to monitor and trade
stocks and to conduct other banking functions), data analysis,
manufacturing, Internet search engines, automobiles, mobile de-
vice applications (e.g., Apple’s Siri speech recognition
software),
aircraft guidance systems, and in a plethora of other
applications
(see Kurzweil, 2005; Russell & Norvig, 2003). Moreover, full
human brain simulation is a possibility in the near future.
Notably,
the Blue Brain Project (Switzerland) aims to create a synthetic
brain by reverse-engineering the mammalian brain down to the
molecular level. In 2009 they successfully developed a model of
rat’s cortex, and a full human brain simulation may be possible
in
20 years (Neild, 2012). In 2013, the Obama administration an-

nounced a billion-dollar investment in a brain mapping project
that
consists of a consortium of both private and public
organizations
(i.e., Defense Applied Research Projects Agency; National
Insti-
tutes for Health, National Science Foundation; Markoff, 2013).
The project aims to create a functional map of neural networks
of
the human brain (see Alivisatos et al, 2012). The current and
planned research and development investment in both the
private
and public sectors are indicative of the focus on the
advancement
of AI and associated technologies. The application of AI
technol-
ogies in the mental health care field is undoubtedly a growth
area
that is destined to have a profound influence on psychological
practice and research in the years ahead.
The field of psychology has historically made important contri-
butions to the field of AI. For example, Frank Rosenblatt was
the
psychologist who built the Mark 1 Perceptron (Rosenblatt,
1957)—the first machine that could learn on its own using
neural
network concepts. The work of neuropsychologist Donald O.
Hebb, whose theory for how neurons learn by the strengthening
of
connections between them (Hebb, 1949), set the foundation for
the
study of artificial neural nets in AI. The work of psychologist
David Rumelhart and colleagues (see Rumelhart, McClelland &
PDP Research Group, 1986) furthered the study of neural-net

models of memory that influenced the development of machine
learning. Moreover, the entire “cognitive revolution” in
psychol-
ogy during the 1960s led to interest in computer models of
human
cognition. The further contributions of psychologists and other
health care professionals in the study, development, and imple-
mentation of AI technology can be expected. Some of the areas
to
which psychologists and others in the mental health care field
may
contribute include research toward the development of new and
creative
approaches to designing AI technologies, laboratory and field
eval-
uation of AI systems, and the study of how humans and AI
interact
with each other. Some other examples of research in this area
may
include study of the social relationships between people and
arti-
ficial intelligent agents as well as the psychological effects of
human-like robots on people (and vice versa). Furthermore, psy-
chologists can contribute to decisions regarding the ethical use
of
this technology in psychological practice, research, and in all
other
areas of society.
As discussed in this article, there are many practical
applications
of AI technology that may serve to benefit patients, health care
providers, and society by enhancing care, increasing efficiency,
and improving access to quality services. There is, nonetheless,
the
risk of this technology having negative implications as well. In

the
near term, specific applied use and collaboration with AI-
enabled
systems that serve to assist mental health care professionals can
be
expected. In the not-so-distant future, the widespread use of the
AI
technologies discussed in this article may be commonplace.
Psy-
chologists and all mental health care professionals must
therefore
be prepared to embrace and guide the use and study of AI tech-
nologies for the benefit of patients, the profession, and society
as
a whole.
References
Alivisatos, A. P., Chun, M., Church, G. M., Greenspan, R. J.,
Roukes,
M. L., & Yuste, R. (2012). The brain activity map project and
the
challenge of functional connectomics. Neuron, 74, 970 –974.
doi:
10.1016/j.neuron.2012.06.006
Anderson, M., & Anderson, S. L. (Eds.). (2011). Machine
ethics. New York,
NY: Cambridge University Press.
doi:10.1017/CBO9780511978036
Asimov, I. (1942). Runaround: Astounding science fiction. New
York, NY:
Street and Smith Publications, Inc.
Asimov, I. (1985). Robots and Empire. New York, NY:

Doubleday.
Bartneck, C., Lyons, M. J., & Saerbeck, M. (2008). The
relationship
between emotion models and artificial intelligence. Proceedings
of the
Workshop on the Role of Emotions in Adaptive Behaviour and
Cognitive
Robotics in affiliation with the 10th International Conference
on Simu-
lation of Adaptive Behavior: From animals to animates (SAB
2008).
Osaka, Japan.
Bindoff, I., Stafford, A., Peterson, G., Kang, B. H., & Tenni, P.
(2012). The
potential for intelligent decision support systems to improve the
quality
and consistency of medication reviews. Journal of Clinical
Pharmacy
and Therapeutics, 37, 452– 458. doi:10.1111/j.1365-
2710.2011.01327.x
Brynjolfsson, E., & McAfee, A. (2011). Race against the
machine: How
the digital revolution is accelerating innovation, driving
productivity,
and irreversibly transforming employment and the economy.
Cambridge,
MA: MIT Sloan School of Management. Retrieved from
http://ebusiness
T
hi
s

do
cu
m
en
t
is
co
py
ri
gh
te
d
by
th
e
A
m
er
ic
an
P
sy
ch
ol

og
ic
al
A
ss
oc
ia
ti
on
or
on
e
of
it
s
al
li
ed
pu
bl
is
he
rs
.

T
hi
s
ar
ti
cl
e
is
in
te
nd
ed
so
le
ly
fo
r
th
e
pe
rs
on
al
us

e
of
th
e
in
di
vi
du
al
us
er
an
d
is
no
t
to
be
di
ss
em
in
at
ed

br
oa
dl
y.
6 LUXTON
.mit.edu/research/Briefs/Brynjolfsson_McAfee_Race_Against_t
he_
Machine.pdf
Buchanan, B. G. (2005). A (very) brief history of artificial
intelligence. AI
Magazine, 26, 53– 60.
Buchanan, B. G., & Shortliffe, E. H. (1984). Rule based expert
systems:
The MYCIN experiments of the Stanford heuristic programming
project.
Reading, MA: Addison Wesley.
Caudell, T. P., & Mizell, D. W. (1992, January). Augmented
reality: An
application of heads-up display technology to manual
manufacturing
processes. In System Sciences, 1992: Proceedings of the twenty-
fifth
Hawaii International Conference on System Sciences (Vol. 2,
pp. 659 –
669). New York, NY: IEEE. doi:10.1109/HICSS.1992.183317
Cortes, C., & Vapnik, V. (1995). Support-vector networks.

Machine Learn-
ing, 20, 273–297. doi:10.1007/BF00994018
Coyle, D., Matthews, M., Sharry, J., Nisbet, A., & Doherty, G.
(2005).
Personal investigator: A therapeutic 3D game for adolescent
psychother-
apy. Interactive Technology and Smart Education, 2, 73– 88.
doi:
10.1108/17415650580000034
DeAngelis, T. (2012, March). A second life for practice?
Monitor on
Psychology, 43. Retrieved from
http://www.apa.org/monitor/2012/03/
avatars.aspx
Defense Applied Research Projects Agency. (2013). Detection
and com-
putational analysis of psychological signals (DCAPS).
Retrieved from
http://www.darpa.mil/Our_Work/I2O/Programs/
Detection_and_Computational_Analysis_of_Psychological_Sign
als-
_(DCAPS).aspx
Finlay, P. N. (1994). Introducing decision support systems.
Cambridge,
MA: Blackwell Publishers.
Fujita, H., & Wu, I.-C. (2012). A special issue on artificial
intelligence in
computer games: AICG. Knowledge-Based Systems, 34, 1–2.
doi:
10.1016/j.knosys.2012.05.014

Garfield, D. A., Rapp, C., & Evens, M. (1992). Natural
language process-
ing in psychiatry: Artificial intelligence technology and
psychopathol-
ogy. Journal of Nervous and Mental Disease, 180, 2227–2237.
Gil, D., & Manuel, D. J. (2009). Diagnosing Parkinson’s by
using artificial
neural networks and support vector machines. Global Journal of
Com-
puter Science and Technology, 9, 63–71.
Gorrindo, T., & Groves, J. (2009). Computer simulation and
virtual reality
in the diagnosis and treatment of psychiatric disorders.
Academic Psy-
chiatry, 33, 413– 417. doi:10.1176/appi.ap.33.5.413
Güzeldere, G., & Franchi, S. (1995). Dialogues with colorful
“personali-
ties” of early AI. Stanford Humanities Review, 4, 161–169.
Hand, D. J. (1985). Artificial intelligence and psychiatry.
Cambridge, UK:
Cambridge University Press.
Hardesty, L. (2012 June 22). Researchers amplify variations in
video,
making the invisible visible. Retrieved from http://web.mit.edu/
newsoffice/2012/amplifying-invisible-video-0622.html
Hardin, J. M., & Chhieng, D. C. (2007). Data mining and
clinical decision
support. In E. S. Berner (Ed.), Clinical decision support
systems: Theory
and practice (2nd ed., pp. 44 – 63). New York, NY: Springer.

doi:
10.1007/978-0-387-38319-4_3
Hebb, D. O. (1949). The organization of behavior. New York,
NY: Wiley.
IBM. (2013). IBM Watson: Ushering in a new era of computing.
Retrieved
from http://www-03.ibm.com/innovation/us/watson/index.shtml
InTouch Health. (2012). RP-VITA robot. Retrieved from
http://www
.intouchhealth.com/products-and-services/products/rp-vita-
robot/
Jagielska, I., Matthews, C., & Whitfort, T. (1999). An
investigation into the
application of neural networks, fuzzy logic, genetic algorithms,
and
rough sets to automated knowledge acquisition for classification
prob-
lems. Neurocomputing, 24, 37–54. doi:10.1016/S0925-
2312(98)00090-3
Kohannim, O., Hua, X., Hibar, D. P., Lee, S., Chou, Y. Y.,
Toga, A. W.,
. . . Thompson, P. M. (2010). Boosting power for clinical trials
using
classifiers based on multiple biomarkers. Neurobiology of
Aging, 31,
1429 –1442. doi:10.1016/j.neurobiolaging.2010.04.022
Kramer, G. M., Mishkind, M. C., Luxton, D. D., & Shore, J. H.
(2013).
Managing risk and protecting privacy in telemental health: An
overview

of legal, regulatory, and risk management issues. In Myers &
Turvey.
(Eds.) Telemental health: Clinical, technical and administrative
foun-
dations for evidence-based practice. New York, NY: Elsevier.
Krijn, M., Emmelkamp, P. M. G., Olafsson, R. P., & Biemond,
R. (2004).
Virtual reality exposure therapy of anxiety disorders: A review.
Clinical
Psychology Review, 24, 259 –281.
doi:10.1016/j.cpr.2004.04.001
Krogh, A. (2008). What are artificial neural networks? Nature
Biotechnol-
ogy, 26, 195–197. doi:10.1038/nbt1386
Kurzweil, R. (2005). The singularity is near. New York, NY:
Viking Press.
Lambert, M. J., & Barley, D. E. (2001). Research summary on
the thera-
peutic relationship and psychotherapy outcome. Psychotherapy:
Theory,
Research, Practice, Training, 38, 357–361. doi:10.1037/0033-
3204.38
.4.357
Lester, J. (2005, January). About Brigadoon. Brigadoon: An
innovative
online community for people dealing with Asperger’s syndrome
and
autism. Retrieved from
http://braintalk.blogs.com/brigadoon/2005/01/
about_brigadoon.html

Linden Research, Inc. (2013). Second Life (Version 1.3.2).
Retrieved from
http://secondlife.com/
Lingley, R., Ali, M., Liao, Y., Mirjalili, R., Klonner, M.,
Sopanen, M.,
. . . Parviz, B. A. (2011). A single-pixel wireless contact lens
display,
Journal of Micromechanics and Microengineering, 21, 125014.
doi:
10.1088/0960-1317/21/12/125014
Manning, C. D., & Schütze, H. (1999). Foundations of
statistical natural
language processing. Cambridge, MA: The MIT Press.
Markoff, J. (2011, March 5). Armies of expensive lawyers,
replaced by
cheaper software. The New York Times. Retrieved from
http://www
.nytimes.com/2011/03/05/science/05legal.html
Markoff, J. (2013, February 18). Obama seeking to boost study
of human
brain. The New York Times. Retrieved from
http://www.nytimes.com/
2013/02/18/science/project-seeks-to-build-map-of-human-
brain.html?
pagewanted�all&_r�0
Masri, R. Y., & Mat Jani, H. (2012, June). Employing artificial
intelligence
techniques in Mental Health Diagnostic Expert System. In
ICCIS 2012:
International Conference on Computer & Information Science
(Vol. 1,

pp. 495– 499). New York, NY: IEEE.
Matthews, M., & Coyle, D. (2010). The role of gaming in
mental health.
In K. Anthony, D. M. Nagel, & S. Goss (Eds.), The use of
technology in
mental health: Applications, ethics and practice (Vol. 40, pp.
134 –142).
Springfield, IL: Charles C. Thomas.
McCarthy, J. (1984). Some expert systems need common sense.
Computer
Culture: The Scientific, Intellectual, and Social Impact of the
Computer,
426, 129 –137.
McShane, M., Beale, S., Nirenburg, S., Jarrell, B., & Fantry, G.
(2012).
Inconsistency as a diagnostic tool in a society of intelligent
agents.
Artificial Intelligence in Medicine, 55, 137–148.
doi:10.1016/j.artmed
.2012.04.005
Minett, J. W., Zheng, H. Y., Manson CM. Fong, Zhou, L., Peng,
G., & SY,
W. (2012). A Chinese text input brain– computer interface
based on the
P300 Speller. International Journal of Human-Computer
Interaction,
28, 472– 483. doi:10.1080/10447318.2011.622970
Moore, G. E. (1965). Cramming more components onto
integrated circuits.
Electronics, 38, 114 –116. doi:10.1109/N-SSC.2006.4785860

Morelli, R. (1989, November). Artificial intelligence in
psychiatry: Issues
and questions. In Proceedings of the annual international
conference of
the IEEE Engineering in Medicine and Biology Society
(EMBS), 1989:
Images of the twenty-first century (pp. 1812–1813). New York,
NY:
IEEE.
Morelli, R. A., Bronzino, J. D., & Goethe, J. W. (1987). Expert
systems in
psychiatry. Journal of Medical Systems, 11, 157–168.
doi:10.1007/
BF00992350
T
hi
s
do
cu
m
en
t
is
co
py
ri
gh
te

d
by
th
e
A
m
er
ic
an
P
sy
ch
ol
og
ic
al
A
ss
oc
ia
ti
on
or
on

e
of
it
s
al
li
ed
pu
bl
is
he
rs
.
T
hi
s
ar
ti
cl
e
is
in
te
nd

ed
so
le
ly
fo
r
th
e
pe
rs
on
al
us
e
of
th
e
in
di
vi
du
al
us
er

an
d
is
no
t
to
be
di
ss
em
in
at
ed
br
oa
dl
y.
PRACTICE
Naam, R. (2010). More than human: Embracing the promise of
biological
enhancement. New York, NY: Broadway Books.

Neild, B. (2012, October 12). Scientists to simulate human brain
inside a
supercomputer. CNN Labs. Retrieved from
http://www.cnn.com/2012/
10/12/tech/human-brain-computer
Patel, V. L., Shortliffe, E. H., Stefanelli, M., Szolovits, P.,
Berthold, M. R.,
Bellazzi, R., & Abu-Hanna, A. (2009). The coming of age of
artificial
intelligence in medicine. Artificial Intelligence in Medicine, 46,
5–17.
doi:10.1016/j.artmed.2008.07.017
Reger, G. M., Holloway, K. M., Rothbaum, B. O., Difede, J.,
Rizzo, A. A.,
& Gahm, G. A. (2011). Effectiveness of virtual reality exposure
therapy
for active duty soldiers in a military mental health clinic.
Journal of
Traumatic Stress, 24, 93–96. doi:10.1002/jts.20574
Riva, G. (2010). Using virtual immersion therapeutically. In K.
Anthony,
D. A. M. Nagel, & S. Goss (Eds.), The use of technology in
mental
health: Applications, ethics and practice (pp. 114 –123).
Springfield, IL:
Charles C Thomas.
Rizzo, A. A., Buckwalter, J. G., & Neumann, U. (1997). Virtual
reality and
cognitive rehabilitation: A brief review of the future. The
Journal of
Head Trauma Rehabilitation, 12, 1–15. doi:10.1097/00001199-
199712000-00002

Rizzo, A. A., Lange, B., Buckwalter, J. G., Forbell, E., Kim, J.,
Sagae, K.,
. . . Kenny, P. (2011). An intelligent virtual human system for
providing
healthcare information and support. Study of Health Technology
Infor-
mation, 163, 503–509.
Rizzo, A. A., Parsons, T. D., Lange, B., Kenny, P., Buckwalter,
J. G.,
Rothbaum, B., . . . Reger, G. (2011). Virtual reality goes to war:
A brief
review of the future of military behavioral healthcare. Journal
of Clin-
ical Psychology in Medical Settings, 18, 176 –187.
doi:10.1007/s10880-
011-9247-2
Rogers, C. (1951). Client-centered therapy. Boston: Houghton
Mifflin
Company.
Rosenblatt, F. (1957), The Perceptron—a perceiving and
recognizing
automaton. Report 85– 460-1, Cornell Aeronautical Laboratory.
Rumelhart, D. E., & McClelland, J. L. (1986). Parallel
distributed pro-
cessing: Explorations in the microstructure of cognition (Vol. 1.
Foun-
dations). Cambridge, MA: MIT Press.
Russell, S. J., & Norvig, P. (2003). Artificial intelligence: A
modern
approach (2nd ed.), Upper Saddle River, NJ: Prentice Hall.

Samuel, A. L. (1959). Some studies in machine learning using
the game of
checkers. Retrieved from
http://www.cs.unm.edu/~terran/downloads/
classes/cs529-s11/papers/samuel_1959_B.pdf
Searle, J. (1980). Minds, brains and programs. Behavioral and
Brain
Sciences, 3, 417– 424. doi:10.1017/S0140525X00005756
Servan-Schreiber, D. (1986). Artificial intelligence and
psychiatry. Journal
of Nervous and Mental Disease, 174, 191–202.
doi:10.1097/00005053-
198604000-00001
Shibata, T., & Wada, K. (2011). Robot therapy: A new approach
for mental
healthcare of the elderly - a mini-review. Gerontology, 57, 378
–386.
doi:10.1159/000319015
Shortliffe, E. H. (1976). Computer-based medical consultations:
MYCIN.
New York, NY: Elsevier.
Shortliffe, E. H. (1993). The adolescence of AI in medicine:
Will the field
come of age in the ‘90s? Artificial Intelligence in Medicine, 5,
93–106.
doi:10.1016/0933-3657(93)90011-Q
Szolovits, P. (1982). Artificial intelligence and medicine.
Boulder, CO:
Westview Press.

Teuscher, C., & Hofstadter, D. R. (2006). Alan Turing: Life and
legacy of
a great thinker. New York, NY: Springer.
Turing, A. M. (1950). Computing machinery and intelligence.
Mind, 49,
433– 460.
Vinge, V. (1993). The coming technological singularity: How to
survive in
the post-human era. Retrieved from http://www-
rohan.sdsu.edu/faculty/
vinge/misc/singularity.html
von Neumann, J. (2012). The computer and the brain (The
Silliman
Memorial Lectures Series). New Haven, CT: Yale University
Press.
Weizenbaum, J. (1966). Computer power and human reason:
From judg-
ment to calculation. San Francisco, CA: Freeman.
Wolpaw, J. R., Birbaumer, N., McFarland, D. J., Pfurtscheller,
G., &
Vaughan, T. M. (2002). Brain-computer interfaces for
communication
and control. Clinical Neurophysiology, 113, 767–791.
doi:10.1016/
S1388-2457(02)00057-3
Zadeh, L. A. (1965). Fuzzy sets. Information and Control, 8,
338 –353.
doi:10.1016/S0019-9958(65)90241-X

Received June 5, 2013
Revision received July 26, 2013
Accepted September 3, 2013 �
T
hi
s
do
cu
m
en
t
is
co
py
ri
gh
te
d
by
th
e
A
m
er

ic
an
P
sy
ch
ol
og
ic
al
A
ss
oc
ia
ti
on
or
on
e
of
it
s
al
li
ed

pu
bl
is
he
rs
.
T
hi
s
ar
ti
cl
e
is
in
te
nd
ed
so
le
ly
fo
r
th

e
pe
rs
on
al
us
e
of
th
e
in
di
vi
du
al
us
er
an
d
is
no
t
to
be

di
ss
em
in
at
ed
br
oa
dl
y.
8 LUXTON
This week we covered Artificial Intelligence. The assignment is
(Discussion post) about 250 -500 words (you don’t have to write
a reference, if you do, please use the one I have uploaded)
What is AI? How has AI been applied in the mental health care
system (provide examples from Luxton, 2013)? What are the
benefits of AI in the health care system, and what are some of
the concerns? What is the goal of Natural Language Processing
(NLP)? Describe each of the main distinct focuses of NLP.
Select two of the levels of NLP and describe each, as well as
provide an example. Describe some of the similarities and
differences between the Statistical Approach and the
Connectionist Approach. How did Fei Fei Li and colleagues
incorporate NLP into their object-naming system described in
the TED Talk? How can the technology described in the TED
Talk be applied to real-world applications?

Natural Language Processing: An Overview

Natural Language Processing: An Overview

Recommended

Recommended

More Related Content

Similar to Natural Language Processing: An Overview

Similar to Natural Language Processing: An Overview (20)

More from deanmtaylor1545

More from deanmtaylor1545 (20)

Recently uploaded

Recently uploaded (20)

Natural Language Processing: An Overview