The document discusses the importance of literacy skills for economic success, noting that too many people lack these skills. It then outlines a federal program that provides discounts to schools and libraries for telecommunications and internet access to promote affordable connectivity. The program allows individual applications as well as applications through consortiums, and requires that user equipment be provided to utilize the funded connectivity resources.
This document describes a space-geosocial application called "LinkAStar" that allows users to search for stars, check-in to stars, and share wishes. The app uses GPS, gyroscope, and star data to allow users to search for stars in the direction their phone is facing. Users can then check-in to a star by adding a comment or wish. The backend uses a star search engine, star rendering engine, and database to manage user check-ins and comments. Future additions may include star recommendations and gamification of check-ins.
The document discusses the importance of literacy skills for economic success, noting that too many people lack these skills. It then outlines a federal program that provides discounts to schools and libraries for telecommunications and internet access to improve literacy. The program allows individual applications or applications through consortiums, and requires that user equipment be provided to utilize the funded connectivity resources.
Robotic pallet assembly can help companies respond to increasing customer demands for custom pallets. Robots allow for high mix production with fast changeovers and consistent quality and throughput. This reduces injuries for employees from dangerous and repetitive tasks. Robotic automation can provide a safer working environment, increased efficiency, and higher profit margins to benefit both the company's financial returns and its employees. Yaskawa Motoman offers robotic pallet assembly solutions and has over 300,000 robots installed globally to support customers locally.
1) A group of 40 people from Camp-In-A-Box went on a mission trip to Haiti in 2010 to help rebuild homes and provide aid to orphanages after the 2010 earthquake.
2) The group set up a camp on a mountain and held activities for local children, including a photo scavenger hunt and talent show.
3) During their time in Haiti, the group helped deliver a woman in labor, provided medical aid, and said emotional goodbyes as their trip came to an end.
The document summarizes a research paper on detecting stable and temporal topics from social media data using a unified mixture model. It proposes a model that distinguishes temporal topics, which are discussed intensely for a short period related to real-world events, from stable topics, which are regularly discussed interests. The model mixes user and temporal features to determine topic type, with stable topics dependent on users and temporal topics dependent on time. An EM algorithm is used to estimate model parameters and maximize the log-likelihood of detecting both topic types from a user-time-associated document collection.
The document discusses Bloom's Taxonomy, which is a classification of learning objectives into different levels of complexity and specificity. It was created by Benjamin Bloom in 1955 to categorize educational goals and objectives. The taxonomy contains three domains: cognitive, affective, and psychomotor. Within the cognitive domain are knowledge, comprehension, application, analysis, synthesis and evaluation. Each level builds on the previous ones as the cognitive process becomes more complex. The document provides examples of verbs to describe each level of cognitive learning. It also discusses the affective and psychomotor domains and provides guidance on how to incorporate Bloom's Taxonomy into teaching practices.
The document discusses transformations of quadratic functions in vertex form f(x) = a(x-h)2 + k. It explains how changing the coefficients a, h, and k affects the graph of the quadratic function. Specifically, it states that changing a widens or narrows the graph, changing h shifts the graph left or right, and changing k shifts the graph up or down. It also provides examples of writing equations for quadratic functions based on given graphs and finding the vertex of a quadratic function in standard form.
This document outlines Oakridge Centre's 50th anniversary marketing campaign. It discusses the mall's history, target markets including suburban affluents and urban professionals, and marketing objectives to maximize customer experience and increase gift card awareness. Recommendations include advertising through magazines, billboards, and transit to promote the anniversary and reinforce the brand image. Direct marketing tactics like mailers and sales promotions are suggested to connect with customers and drive traffic. Public relations efforts will focus on sharing customer stories and benchmarking the campaign's success.
The Ingenio programme is a voluntary and free entrepreneurship programme at Deusto University that aims to foster an innovative and entrepreneurial spirit among final-year students. The programme looks for willing and positive students who enjoy taking initiatives and are enthusiastic, optimistic, and eager to actively participate, commit to the group, and generate new ideas. It uses workshops, seminars, and other training to develop competencies like communication, team building, creativity, self-motivation, and turning ideas into business plans.
This document outlines the process for planning video lessons, which includes identifying student objectives and assessments, determining the unit essential question and concepts to be taught, developing lesson essential questions, recording and developing the video lesson, and assessing student progress. The process emphasizes understanding the key topics and concepts students should understand, and developing lessons and assessments around those ideas.
The document discusses the importance of literacy skills for economic success, noting that too many people lack these skills. It then outlines a federal program that provides discounts to schools and libraries for telecommunications and internet access to promote affordable connectivity. The program allows individual applications as well as applications through consortiums, and requires that user equipment be provided to utilize the funded connectivity resources.
This document describes a space-geosocial application called "LinkAStar" that allows users to search for stars, check-in to stars, and share wishes. The app uses GPS, gyroscope, and star data to allow users to search for stars in the direction their phone is facing. Users can then check-in to a star by adding a comment or wish. The backend uses a star search engine, star rendering engine, and database to manage user check-ins and comments. Future additions may include star recommendations and gamification of check-ins.
The document discusses the importance of literacy skills for economic success, noting that too many people lack these skills. It then outlines a federal program that provides discounts to schools and libraries for telecommunications and internet access to improve literacy. The program allows individual applications or applications through consortiums, and requires that user equipment be provided to utilize the funded connectivity resources.
Robotic pallet assembly can help companies respond to increasing customer demands for custom pallets. Robots allow for high mix production with fast changeovers and consistent quality and throughput. This reduces injuries for employees from dangerous and repetitive tasks. Robotic automation can provide a safer working environment, increased efficiency, and higher profit margins to benefit both the company's financial returns and its employees. Yaskawa Motoman offers robotic pallet assembly solutions and has over 300,000 robots installed globally to support customers locally.
1) A group of 40 people from Camp-In-A-Box went on a mission trip to Haiti in 2010 to help rebuild homes and provide aid to orphanages after the 2010 earthquake.
2) The group set up a camp on a mountain and held activities for local children, including a photo scavenger hunt and talent show.
3) During their time in Haiti, the group helped deliver a woman in labor, provided medical aid, and said emotional goodbyes as their trip came to an end.
The document summarizes a research paper on detecting stable and temporal topics from social media data using a unified mixture model. It proposes a model that distinguishes temporal topics, which are discussed intensely for a short period related to real-world events, from stable topics, which are regularly discussed interests. The model mixes user and temporal features to determine topic type, with stable topics dependent on users and temporal topics dependent on time. An EM algorithm is used to estimate model parameters and maximize the log-likelihood of detecting both topic types from a user-time-associated document collection.
The document discusses Bloom's Taxonomy, which is a classification of learning objectives into different levels of complexity and specificity. It was created by Benjamin Bloom in 1955 to categorize educational goals and objectives. The taxonomy contains three domains: cognitive, affective, and psychomotor. Within the cognitive domain are knowledge, comprehension, application, analysis, synthesis and evaluation. Each level builds on the previous ones as the cognitive process becomes more complex. The document provides examples of verbs to describe each level of cognitive learning. It also discusses the affective and psychomotor domains and provides guidance on how to incorporate Bloom's Taxonomy into teaching practices.
The document discusses transformations of quadratic functions in vertex form f(x) = a(x-h)2 + k. It explains how changing the coefficients a, h, and k affects the graph of the quadratic function. Specifically, it states that changing a widens or narrows the graph, changing h shifts the graph left or right, and changing k shifts the graph up or down. It also provides examples of writing equations for quadratic functions based on given graphs and finding the vertex of a quadratic function in standard form.
This document outlines Oakridge Centre's 50th anniversary marketing campaign. It discusses the mall's history, target markets including suburban affluents and urban professionals, and marketing objectives to maximize customer experience and increase gift card awareness. Recommendations include advertising through magazines, billboards, and transit to promote the anniversary and reinforce the brand image. Direct marketing tactics like mailers and sales promotions are suggested to connect with customers and drive traffic. Public relations efforts will focus on sharing customer stories and benchmarking the campaign's success.
The Ingenio programme is a voluntary and free entrepreneurship programme at Deusto University that aims to foster an innovative and entrepreneurial spirit among final-year students. The programme looks for willing and positive students who enjoy taking initiatives and are enthusiastic, optimistic, and eager to actively participate, commit to the group, and generate new ideas. It uses workshops, seminars, and other training to develop competencies like communication, team building, creativity, self-motivation, and turning ideas into business plans.
This document outlines the process for planning video lessons, which includes identifying student objectives and assessments, determining the unit essential question and concepts to be taught, developing lesson essential questions, recording and developing the video lesson, and assessing student progress. The process emphasizes understanding the key topics and concepts students should understand, and developing lessons and assessments around those ideas.
This document discusses improving lighting in a parking structure and proposes a new lighting system called LimeLight. It describes how LimeLight uses high output fluorescent fixtures, wireless controls, and smart technology to provide energy savings, increase safety and maintenance efficiencies, and deliver information-rich reporting. The system is presented as having a 3-year ROI of 65% in energy cost reduction and providing active lighting, maintenance benefits, and optimal long-term support.
This document summarizes training services offered by an organization for industry professionals. It provides details on standard and customized training programs in CAD, CAE, and engineering domains, as well as communicational and leadership skills. The organization has experienced faculty and domain experts, conducts training at their premises or client locations, and has partnerships to offer ASME certified programs. It also lists some client companies and certified engineers that were placed in industry.
Cutting removes information from its original location, while copying leaves the information in the original location and makes additional copies. Information cut or copied remains on the clipboard until replaced, and can be pasted an infinite number of times. To cut, copy, or paste, you can select text and use menu options or right-click, or use keyboard shortcuts like Ctrl-C to copy and Ctrl-V to paste.
文献紹介:Token Shift Transformer for Video ClassificationToru Tamaki
Hao Zhang, Yanbin Hao, Chong-Wah Ngo, Token Shift Transformer for Video Classification, ACM MM '21: Proceedings of the 29th ACM International Conference on MultimediaOctober 2021 Pages 917–925https://doi.org/10.1145/3474085.3475272
http://vireo.cs.cityu.edu.hk/papers/Hao_MM2021.pdf
http://arxiv.org/abs/2108.02432
https://dl.acm.org/doi/abs/10.1145/3474085.3475272
文献紹介:Selective Feature Compression for Efficient Activity Recognition InferenceToru Tamaki
Chunhui Liu, Xinyu Li, Hao Chen, Davide Modolo, Joseph Tighe; Selective Feature Compression for Efficient Activity Recognition Inference, Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), 2021, pp. 13628-13637
https://openaccess.thecvf.com/content/ICCV2021/html/Liu_Selective_Feature_Compression_for_Efficient_Activity_Recognition_Inference_ICCV_2021_paper.html
文献紹介:An Image is Worth 16x16 Words: Transformers for Image Recognition at ScaleToru Tamaki
Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, Neil Houlsby, An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale, ICLR2021.
https://openreview.net/forum?id=YicbFdNTTy
1. SIGIR 2013
(Users and Interactive IR I)
デンソーアイティーラボラトリ 山本光穂
資料中の図は論文を引用しております。
13年10月9日水曜日
2. 資料中の図は論文を引用しております。
An Effective Implicit Relevance Feedback Technique
Using Affective, Physiological and Behavioural Features
• Implicit Relevance Feedbackとは?
• ユーザの意図を推定した上で、関連する文章や結果を提示する
ことによって検索結果品質を向上させる手法[1]
• ユーザの意図を推定するためのfeatureとして以下の値を使えば
良い結果が得られることが知られている[2]
• dwell time: あるドキュメントに対する滞在時間
• task intension: 検索意図
• task intensionについてはユーザに直接聞くことはできない事
から、何らかの値を使って推定する必要
2
[1] Accurately interpreting clickthrough data as implicit feedback (SIGIR 2005) Thorsten Joachims , et all.
[2] A study on the effects of personalization and task information on implicit feedback performance(CIKM
2006) Ryen W. White, et all.
13年10月9日水曜日
4. 資料中の図は論文を引用しております。
検証の流れ-1
1.被験者に対してvideo retrieval systemを利用させ、以下
の4つの検索タスクを実行させる
• INS Task (Information seeking intent)
• INF Task (re-finding search intent)
• ENA Task (entertainment-based search intent where searchers
adjust their arousal level)
• ENM Task (the entertainment- based search intent where searchers
adjust their mood
• なお、各タスク実施時には
あるタスクのお題が出されます。
•
4
BA
Figure 1: A snapshot of the video retrieval system for query “avengers”.
nology into consumer and industrial end-applications. Neu-
roSky MindKit-EMTM
features two key technologies: (i)
ThinkGear-EMTM
headset and (ii) eSense-EMTM
software
(i.e. brainwave interpretation software). The headset is
used to extract, filter, and amplify brainwave (EEG) signals
and convert that information into digital mental state out-
puts for eSense-EMTM
software. The EEG signals read by
the MindKit-EMTM
are detected on the forehead via points
Fp1 (electrode placement system by the International Fed-
eration in Encephalography and Clinical Neurophysiology).
The headset has three dry active sensors: one sensor located
on the forehead and two sensors are located behind the ears
as ground/reference sensors. It also has electronic circuitry
that filters and amplifies the brainwaves. The eSense-EMTM
software further processes and analyses the obtained brain-
wave signals into two useful neuro-sensory values: the user’s
Attention4
and Meditation5
levels at any given moment.
The output of eSense-EMTM
software has been tested over
a wide population and under di↵erent environmental condi-
the output of the BodyMedia SenseWearR
Pro3 Armband;
and the Attention or Meditation data (referred to as “NV”)
from the output of the eSense-EMTM
software. For our be-
havioural signal, we considered the dwell time (referred to
as “DT”) logged by the system as our dwell time feature.
Finally the task intention was considered as task feature
(referred to as “Task”).
Preprocessing: For each visited video, the value of
each sensory feature (for both a↵ective and physiological
features) was calculated by averaging the data logged by
its sensory device during the dwell time period. Since none
of the instruments we used normalised the data, we scaled
signal values before applying any classification method, to
avoid having attributes in greater numeric ranges dominat-
ing those in smaller numeric ranges.
3.4 Video Retrieval System
For the completion of the search tasks we used a custom-
made search environment (named VideoHunt) that was de-13年10月9日水曜日
7. 資料中の図は論文を引用しております。
各タスクの詳細とユーザへのお題-2
3)ENA Task(Entertainment-based search intent where
searchers adjust their arousal level)
• 楽しみ・快楽の為にある情報を検索する。
• 覚醒レベルを上げるため (眠気を取り除くため)
4)ENM Task(the entertainment-based search intent
where searchers adjust their mood)
• 楽しみ・快楽の為にある情報を検索する。
• 現状のムードを変更するために映像をみる。
7
あなたが工場で夜警をしていることを想像してください。あなたはちょうど工場内のチェックが終了し、次のチェック
まで時間が少しだけあるとします。あなたは少しつかれていますが、次の回のチェックまで気分転換の為になにかビデ
オをみようと決めました。
あなたは彼氏/彼女と旅行しているとします。通常あなたたちは何らかの理由からめったにあえないとします。旅行期間
があと数日になり、彼氏/彼女はあなたと会えなくなることを寂しく思っています。このような気持ちを変えるためにあ
なたはなにかビデオを見ようときめました。
13年10月9日水曜日
8. • affective,physiologicalの情報を組み合わせるとランダム値と比較して推定精度が5%程度改善
• さらに、Dwell Timeと組み合わせると15%くらい精度が改善
資料中の図は論文を引用しております。
結果(ユーザの検索意図の推定精度)
8
Table 2: This table shows the prediction accuracy of a model trained on di↵erent sets of features (presented as rows), given di↵erent se
ntentions (presented as columns). The best performing set of features for each condition and search intention is highlighted in bold.
INS ENA ENM INF ALL - INF ALL
Random [BL1](*) 54.88% 64.06% 64.53% 98.83% 61.19% 50.60%
DT [BL2](†) 62.40% 65.62% 66.16% 98.83% 71.31% 72.74%
DT+Task [BL3](‡) –% –% –% –% 69.65% 76.63%
FX 55.63%*
(+1.3%)
66.4%**
(+3.6%)
64.53%
(+0%)
98.83%
(+0%)
62.43%**
(+2.0%)
64.54%**
(+27.4%)
AB 54.88%
(+0%)
64.06%
(+0%)
64.53%
(+0%)
98.83%
(+0%)
61.19%
(+0%)
50.60%
(+0%)
HR 57.89%**
(+5.4%)
64.06%
(+0%)
64.53%
(+0%)
98.83%
(+0%)
62.93%**
(+2.8%)
53.27%**
(+5.2%)
NV 55.63%*
(+1.3%)
64.06%
(+0%)
64.53%
(+0%)
98.83%
(+0%)
61.19%
(+0%)
55.73%**
(+10.1%)
FX+AB+HR+NV 55.63%*
(+1.3%)
69.53%**
(+8.5%)
64.53%
(+0%)
98.83%
(+0%)
67.16%**
(+9.7%)
65.98%**
(+30.3%)
DT+FX 67.66%††
(+8.4%)
68.75%††
(+4.7%)
71.63%††
(+8.2%)
98.83%
(+0%)
72.88%††
(+2.2%)
77.04%††
(+5.9%)
DT+AB 66.91%††
(+7.2%)
67.96%††
(+3.5%)
81.56%††
(+23.2%)
98.83%
(+0%)
71.64%
(+0.4%)
76.22%††
(+4.5%)
DT+HR 63.15%†
(+1.2%)
73.43%††
(+11.9%)
82.26%††
(+24.3%)
98.83%
(+0%)
72.13%†
(+1.1%)
76.22%††
(+4.5%)
DT+NV 64.41%††
(+3.2%)
70.31%††
(+7.1%)
74.46%††
(+12.5%)
98.83%
(+0%)
72.13%†
(+1.1%)
75.40%††
(+5.6%)
DT+FX+AB+HR+NV 66.16%††
(+6.0%)
75%††
(+14.2%)
80.14%††
(+21.1%)
98.83%
(+0%)
75.37%††
(+5.6%)
77.04%††
(+5.9%)
DT+Task+FX+AB+HR+NV –% –% –% –% 76.36%‡‡
(+9.6%)
78.89%‡‡
(+2.9%)
the prediction accuracy of a model trained on dwell time and
task features significantly (i.e. “DT+Task”row). The results
also show that the prediction accuracy of such a model is
even higher than a model trained on all features except task
one (i.e. “DT+FX+AB+HR+NV” row). This show that re-
searches on task prediction are complementary to this study
rather than contradictory.
An interesting finding is that the discriminative powe
sensory signals changes once they are combined with d
time, even though they show no such power individually.
example, a sensory feature that was not discriminative
its own for a task (e.g. “HR” feature for “ENM” task), w
combined with dwell time, can result in the highest pre
tion accuracy (i.e. “DT+HR” features for “ENM” task)13年10月9日水曜日
10. 資料中の図は論文を引用しております。
実験手法
• ユーザに特定のテーマについて検索させ、検索クエリー(qv),システムの認識
結果(qtr),ユーザの検索結果の選択履歴を取得
10
①音声クエリを入力 ②認識結果を提示 ③検索結果を表示。ユーザは
求めていた情報の場合クリック
indicate its various statuses, which includes: starting or stopping
“listening” a voice query; displaying the transcribed query; and
failing to generate the transcribed query. These audio cues are
very useful in our transcriptions of the experiment recordings.
(a) (b) (c)
Figure 1. Screenshots of the Google search app on iPad.
3.2 Search Tasks and Topics
Our experiment setting is similar to the one adopted by the
TREC session track [17], in which users can issue multiple
queries to work on one search topic.13年10月9日水曜日
11. 資料中の図は論文を引用しております。
How Do Users Respond to Voice Input Errors?
Lexical and Phonetic Query Reformulation in Voice
• 検索タスク
• TREC 50 task(30 form robust track, 20 from web track)
• 検索時間
• 二分間
• ユーザがして良いこと
• クエリの再構成
• googleのクエリサジェスチョンを利用すること
• 検索結果のブラウジング及びクリック
• 以上を20人のネイティブ・スピーカーに対して実験
11
13年10月9日水曜日
12. 資料中の図は論文を引用しております。
実験の流れ
12
EXPERIMENT PROCEDURE (90 MIN)
User
Background
Questionnaire
Training
(One TREC Topic)
(10 Topics) Interview
10 min
Break
(15 Topics)
Work on a TREC
topic for 2 min
Post-task
questionnaire
12
13年10月9日水曜日
13. 資料中の図は論文を引用しております。
• 908のクエリにエラーを含む (55% of 1650)
• 810がクエリの認識エラー
• 98が不適切なシステムの割り込み
• 単語が認識できなかった場合は検索精度にそれほど影響を与えな
い。
• 一方で、誤認識した場合は検索精度に多大な影響を与える。
13
• 908 queries have voice input errors (55% of 1,650)
• 810 by speech recognition error
• 98 by improper system interruption
45%
49%
6%
% of all 1,650 voice queries
No Error
Speech Rec Error
Improper System
Interruption
1
QUERY TRANSCRIPTION
• qv (a voice query’s actual content)
• manually transcribed from the recording
• two authors had an agreement of 100%, except on
casing, plurals, and prepositions
• qtr (the system’s transcription of a voice query)
• available from the log
16
QUERY TRANSCRIPTION
• qv (a voice query’s actual content)
• manually transcribed from the recording
• two authors had an agreement of 100%, except on
casing, plurals, and prepositions
• qtr (the system’s transcription of a voice query)
• available from the log
16
INDIVIDUAL QUERIES: WORDS
• Missing words: words in qv but not in qtr
• Incorrect words: words in qtr but not in qv
qv: a voice query’s
actual content
qtr: the system’s
transcription
missing
words
incorrect
words
20
INDIVIDUAL QUERIES: PERFORMANCE
• Significant decline of search performance (nDCG@10)
No Errors
742 Queries
Speech Rec Errors
810 Queries
mean SD mean SD
nDCG@10 of qv
0.275 0.20 0.264 0.22
nDCG@10 of qtr
0.275 0.20 0.083 0.16
nDCG@10 - - -0.182 0.23
23
結果1 クエリのエラー率と検索精度
ユーザのクエリの修正方法
13年10月9日水曜日
14. 資料中の図は論文を引用しております。
結果2 ユーザのクエリの修正方法
• クエリの追加(ADD)
• /
• クエリの言い換え(SUB)
14
TEXTUAL PATTERNS
• Query Term Addition (ADD)
• Query Term Substitution (SUB)
• SUB word pairs are manually coded (93% agreement)
Voice Query Transcribed Query ADD words
q1 the sun the son
q2 the sun solar system the sun solar system solar system
Voice Query Transcribed Query SUB words
q1 art theft test
q2 art embezzlement are in Dublin theft embezzlement
q3 stolen artwork stolen artwork embezzlement stolen
art artwork
TEXTUAL PATTERNS
• Query Term Addition (ADD)
• Query Term Substitution (SUB)
• SUB word pairs are manually coded (93% agreement)
Voice Query Transcribed Query ADD words
q1 the sun the son
q2 the sun solar system the sun solar system solar system
Voice Query Transcribed Query SUB words
q1 art theft test
q2 art embezzlement are in Dublin theft embezzlement
q3 stolen artwork stolen artwork embezzlement stolen
art artwork
33
クエリを修正する理由は(1)音声入力の認識結果の修正/(2)検索結果修正
の二点
13年10月9日水曜日
15. 資料中の図は論文を引用しております。
結果2 ユーザのクエリの修正方法
• クエリの削除(RMV)
• /
• クエリの順序の変更(ORD)
15
TEXTUAL PATTERNS
• Query Term Removal (RMV)
• Query Term Reordering (ORD)
Voice Query Transcribed Query
q1 advantages of same sex schools andy just open it goes
q2 same sex schools same sex schools
Voice Query Transcribed Query
q1 interruptions to ireland peace talk is directions to ireland peace talks
q2 ireland peace talk interruptions ireland peace talks interruptions
34
TEXTUAL PATTERNS
• Query Term Removal (RMV)
• Query Term Reordering (ORD)
Voice Query Transcribed Query
q1 advantages of same sex schools andy just open it goes
q2 same sex schools same sex schools
Voice Query Transcribed Query
q1 interruptions to ireland peace talk is directions to ireland peace talks
q2 ireland peace talk interruptions ireland peace talks interruptions
34
13年10月9日水曜日
16. 資料中の図は論文を引用しております。
結果2 ユーザのクエリの修正方法
• ・
16
エラー後は,クエリの順序を変えたり,
クエリ中の単語を削除したりして次のクエリを入力する傾向
• When previous query has voice input error
• Increased use of SUB & ORD
• Less use of ADD & RMV
Patterns Prev Q Error Prev Q No Error Overall
ADD 90.50% 32.98% 53.82%
SUB 15.04% 16.34% 14.87%
RMV 66.75% 37.93% 48.37%
ORD 33.51% 43.03% 39.58%
(All Lexical) 99.74% 77.36% 85.47%
37
etic patterns are nearly always
with previous voice input errors
Prev Q Error Prev Q No Error Overall
0% 14.84% 9.46%
0% 0.60% 0.39%
0% 0.90% 0.57%
0.26% 9.30% 6.02%
0.26% 25.64% 16.44%
0% 20.54% 13.58%
38
• Use of phonetic patterns are nearly always
associated with previous voice input errors
Patterns Prev Q Error Prev Q No Error O
STR/SLW 0% 14.84% 9
SPL 0% 0.60% 0
DIF 0% 0.90% 0
WE 0.26% 9.30% 6
(All Phonetic) 0.26% 25.64% 1
Repeat 0% 20.54% 1
13年10月9日水曜日
17. 資料中の図は論文を引用しております。
結果2 ユーザのクエリの修正方法(発話)
• WE
• すべてのクエリを強く言い直す
• REP
• リピートする
17
PHONETIC PATTERNS
• Partial Emphasis (PE)
• Overstate a specific part of a query
PE Type Example Explanation
Stressing (STR) rap and crime put stress on “rap”
Slow down (SLW) rap and c-r-i-m-e slow down at “crime”
Spelling (SPL) P·u·e·r·t·o Rico spell out each letter in “Puerto”
Different
Pronunciation (DIF)
Puerto Rico pronounce “Puerto” differently
35
13年10月9日水曜日
18. 資料中の図は論文を引用しております。
結果-2 検索結果の修正手法
18
クエリの間違いが事前にあった場合は、クエリの修正を発話によって行うユー
ザはいる。
(ただ効果はいまいちらしい→音声認識は標準的な発話を前提としているため)
• Use of phonetic patterns are nearly always
associated with previous voice input errors
Patterns Prev Q Error Prev Q No Error Overall
STR/SLW 0% 14.84% 9.46%
SPL 0% 0.60% 0.39%
DIF 0% 0.90% 0.57%
WE 0.26% 9.30% 6.02%
(All Phonetic) 0.26% 25.64% 16.44%
Repeat 0% 20.54% 13.58%
38
onetic patterns are nearly always
d with previous voice input errors
Prev Q Error Prev Q No Error Overall
0% 14.84% 9.46%
0% 0.60% 0.39%
0% 0.90% 0.57%
0.26% 9.30% 6.02%
) 0.26% 25.64% 16.44%
0% 20.54% 13.58%
38
• Use of phonetic patterns are nearly alwa
associated with previous voice input err
Patterns Prev Q Error Prev Q No Error
STR/SLW 0% 14.84%
SPL 0% 0.60%
DIF 0% 0.90%
WE 0.26% 9.30%
(All Phonetic) 0.26% 25.64%
Repeat 0% 20.54%
13年10月9日水曜日
19. 資料中の図は論文を引用しております。
結果-3 クエリ修正を利用した場合の検索精度
19
若干精度が向上するとのこと。
• Overall slightly improvement (10% in nDCG@10)
• But highly depends on whether or not voice input
error happened after query reformulation
• Did not reduce the likelihood of voice input errors
The reformulated
query has / is
nDCG@10
(before after)
# of cases
No Error 0.150 0.233 474 (40%)
Speech Rec Error 0.104 0.079 597 (51%)
Interruption 0.156 0.056 79 (6.7%)
Query Suggestion 0.201 0.223 32 (2.7%)
Overall 0.129 0.143 1,182
40
13年10月9日水曜日
20. 資料中の図は論文を引用しております。
Mining Touch Interaction Data on Mobile Devices to
Predict Web
• タッチデバイス環境下でのビヘイビアログを使った,検索結果の適合
性予測
• 調べたい事
• ユーザは、マウスやキーボードをデスクトップコンピュータを使用する場合と比較
してタッチ対応のモバイルデバイスで異なる検索結果文書を表示するか否か
• 関連の低いドキュメントと比較して関連性の高い検索結果文書を表示した場合、タ
ッチ対応モバイルデバイス上で動作が異なるか否か
• 実験手法
• 以下のユーザに対して実験を実施
• smartphoneを使うユーザ:26人
• desktop pcを使うユーザ:30人
20
13年10月9日水曜日