The wide development of mobile applications provides a considerable amount of data of all types (images, texts, sounds, videos, etc.). In this sense, Mobile Context-aware Recommender Systems (MCRS) suggest the user suitable information depending on her/his situation and interests. Two key questions have to be considered 1) how to recommend the user information that follows his/her interests evolution? 2) how to model the user’s situation and its related interests? To the best of our knowledge, no existing work proposing a MCRS tries to answer both questions as we do. This paper describes an ongoing work on the implementation of a MCRS based on the hybrid-ε-greedy algorithm we propose, which combines the standard ε-greedy algorithm and both content-based filtering and case-based reasoning techniques.
Proactive Intelligent Home System Using Contextual Information and Neural Net...IJERA Editor
Nowadays, cities around the world intend to use information technology to improve the lives of their citizens.
Future smart cities will incorporate digital data and technology to interact differently with their human
inhabitants.
Among the key component of a smart city, we find the smart home component. It is an autonomic environment
that can provide various smart services by considering the user’s context information. Several methods are used
in context-aware system to provide such services. In this paper, we propose an approach to offer the most
relevant services to the user according to any significant change of his context environment. The proposed
approach is based on the use of context history information together with user profiling and machine learning
techniques. Experimentations show that the proposed solution can efficiently provide the most useful services to
the user in an intelligent home environment.
DYNAMIC AND REALTIME MODELLING OF UBIQUITOUS INTERACTIONcscpconf
Ubiquitous systems require user to be dynamically and realtime informed in order to make his current activity increasingly easy. First, this paper presents and discusses a method to model the realtime interaction of the user with a ubiquitous system based on Petri-nets modelling technology. The goal deals with investigating dynamically the appropriate form of interaction depending on the context of the user. Thus, the interaction model structure should be dynamically improved with respect to the current and particular activity or goal of the user to better cope with his runtime requirements. This mechanism has been characterized as “models mutation”. Secondly, this paper proves the dynamic construction of models while basing on the dynamic composition of services. The ultimate purpose is to take advantage of the ontology of service written in OWL-S in order to describe the dynamic aspect of Petri-nets based models, especially, the realtime and automatic composition of such models. Simulation work has been conducted to validate the proposed approach.
Conceptual Design of Fuzzy Rule Based Context Aware Meeting Room SystemEditor IJMTER
Due to the exponential growth of wireless network based miniaturized device and
sensors, the dream of context aware ubiquitous computing world is becoming realistic. In this
ubiquitous world of context aware applications the users can get the information and share the
information elsewhere instantaneously. Context aware meeting room is one of the interesting context
aware systems where in the meeting of a given set of users will be organized as per the situation of
users. In this paper we present the conceptual design and development of service recommendation
system for prototypical context aware meeting room using Fuzzy Rules .The proposed
recommendation system for context aware meeting room recommends the services by considering
the users contextual parameters like role, priority and environmental conditions. The fuzzy rule is
constructed using history database and knowledge base of the meeting.
Learning Process Interaction Aided by an Adapter Agentpaperpublications3
Abstract: Computational models have played an important role in the discovery and understanding of the complexities during the learning process. One complexity is the distraction factor on educator-learner interaction affecting the quality of the learning process.
We model an adaptive system model able to dynamically adapt considering user’s performance, simulating the learner as a museum user and the educator as an exhibition module using BDI agents; we adapt the BDI architecture using Type-2 fuzzy inference system to add perceptual human-like capability on agents in order to describe the interaction on user's experience. The resulting model allows content adaptation by creating a personalized interaction environment.
This poster introduces an experimental application to demonstrate the usage of context-based awareness cues in status updates, especially in SNS’s (Social Networking Services). The presented application allows users to add different descriptions of context information to their Twitter messages and Facebook status updates in a narrative format.
RoutineMaker: Towards End-User Automation of Daily Routines Using SmartphonesVille Antila
People use smartphones in daily activities for accessing and storing information in various situations. In this paper, we present a work in progress for detecting and automating some of these activities. To explore the possible patterns we developed an experimental application to detect daily tasks used by smartphones and analyzed it to provide suggestions for “routines”. We conducted a two-week user study with 10 users to evaluate the approach. During the study the application logged the usage patterns, sent information to the server where it was analysed and clustered. The participants could also automate their smartphone tasks using the analysed data. The findings suggest that people would be willing to automatize tasks given that the approach gives flexibility and expressiveness without too much information overload. Future work includes refining the algorithms based on the gathered real-life data and modifying the interaction design to approach the challenges found with the initial study.
Review and analysis of machine learning and soft computing approaches for use...IJwest
The adequacy of user models depends mainly on the accuracy and precision of information that is retrieved to the user. The real challenge in user modelling studies is due to the inadequacy of data, improper use of techniques, noise within the data and imprecise nature of human behavior. For the best results of user modelling, one should choose an appropriate way to do it i.e. by selecting the best suitable approach for the desired domain. Machine learning and Soft computing Techniques have the ability to handle the uncertainty and are extensively being used for user modeling purpose. This paper reviews various approaches of user modeling and critically analyzes the machine learning and soft computing techniques that have successfully captured and formally modelled the human behavior.
Understanding the Privacy Implications of Using Context-based Awareness Cues ...Ville Antila
Information from the physical world is increasingly being digitalized and shared in social networks. We share our locations, tag photos and add different kinds of informal awareness cues about the physical world to our online communities. In this paper, we investigate the privacy implications of shared context cues in social networking services. We present an experimental mobile application, which allows users to add different descriptions of context information to their Facebook and Twitter status updates. The application was used by 12 persons during a two-week user trial using their own devices and Facebook accounts. The results indicate that user-defined abstractions of context items were often preferred over more accurate indicators due to privacy concerns or discomfort in sharing. We also found out that using shared context from friends in vicinity needs careful design to overcome the extended privacy implications.
Proactive Intelligent Home System Using Contextual Information and Neural Net...IJERA Editor
Nowadays, cities around the world intend to use information technology to improve the lives of their citizens.
Future smart cities will incorporate digital data and technology to interact differently with their human
inhabitants.
Among the key component of a smart city, we find the smart home component. It is an autonomic environment
that can provide various smart services by considering the user’s context information. Several methods are used
in context-aware system to provide such services. In this paper, we propose an approach to offer the most
relevant services to the user according to any significant change of his context environment. The proposed
approach is based on the use of context history information together with user profiling and machine learning
techniques. Experimentations show that the proposed solution can efficiently provide the most useful services to
the user in an intelligent home environment.
DYNAMIC AND REALTIME MODELLING OF UBIQUITOUS INTERACTIONcscpconf
Ubiquitous systems require user to be dynamically and realtime informed in order to make his current activity increasingly easy. First, this paper presents and discusses a method to model the realtime interaction of the user with a ubiquitous system based on Petri-nets modelling technology. The goal deals with investigating dynamically the appropriate form of interaction depending on the context of the user. Thus, the interaction model structure should be dynamically improved with respect to the current and particular activity or goal of the user to better cope with his runtime requirements. This mechanism has been characterized as “models mutation”. Secondly, this paper proves the dynamic construction of models while basing on the dynamic composition of services. The ultimate purpose is to take advantage of the ontology of service written in OWL-S in order to describe the dynamic aspect of Petri-nets based models, especially, the realtime and automatic composition of such models. Simulation work has been conducted to validate the proposed approach.
Conceptual Design of Fuzzy Rule Based Context Aware Meeting Room SystemEditor IJMTER
Due to the exponential growth of wireless network based miniaturized device and
sensors, the dream of context aware ubiquitous computing world is becoming realistic. In this
ubiquitous world of context aware applications the users can get the information and share the
information elsewhere instantaneously. Context aware meeting room is one of the interesting context
aware systems where in the meeting of a given set of users will be organized as per the situation of
users. In this paper we present the conceptual design and development of service recommendation
system for prototypical context aware meeting room using Fuzzy Rules .The proposed
recommendation system for context aware meeting room recommends the services by considering
the users contextual parameters like role, priority and environmental conditions. The fuzzy rule is
constructed using history database and knowledge base of the meeting.
Learning Process Interaction Aided by an Adapter Agentpaperpublications3
Abstract: Computational models have played an important role in the discovery and understanding of the complexities during the learning process. One complexity is the distraction factor on educator-learner interaction affecting the quality of the learning process.
We model an adaptive system model able to dynamically adapt considering user’s performance, simulating the learner as a museum user and the educator as an exhibition module using BDI agents; we adapt the BDI architecture using Type-2 fuzzy inference system to add perceptual human-like capability on agents in order to describe the interaction on user's experience. The resulting model allows content adaptation by creating a personalized interaction environment.
This poster introduces an experimental application to demonstrate the usage of context-based awareness cues in status updates, especially in SNS’s (Social Networking Services). The presented application allows users to add different descriptions of context information to their Twitter messages and Facebook status updates in a narrative format.
RoutineMaker: Towards End-User Automation of Daily Routines Using SmartphonesVille Antila
People use smartphones in daily activities for accessing and storing information in various situations. In this paper, we present a work in progress for detecting and automating some of these activities. To explore the possible patterns we developed an experimental application to detect daily tasks used by smartphones and analyzed it to provide suggestions for “routines”. We conducted a two-week user study with 10 users to evaluate the approach. During the study the application logged the usage patterns, sent information to the server where it was analysed and clustered. The participants could also automate their smartphone tasks using the analysed data. The findings suggest that people would be willing to automatize tasks given that the approach gives flexibility and expressiveness without too much information overload. Future work includes refining the algorithms based on the gathered real-life data and modifying the interaction design to approach the challenges found with the initial study.
Review and analysis of machine learning and soft computing approaches for use...IJwest
The adequacy of user models depends mainly on the accuracy and precision of information that is retrieved to the user. The real challenge in user modelling studies is due to the inadequacy of data, improper use of techniques, noise within the data and imprecise nature of human behavior. For the best results of user modelling, one should choose an appropriate way to do it i.e. by selecting the best suitable approach for the desired domain. Machine learning and Soft computing Techniques have the ability to handle the uncertainty and are extensively being used for user modeling purpose. This paper reviews various approaches of user modeling and critically analyzes the machine learning and soft computing techniques that have successfully captured and formally modelled the human behavior.
Understanding the Privacy Implications of Using Context-based Awareness Cues ...Ville Antila
Information from the physical world is increasingly being digitalized and shared in social networks. We share our locations, tag photos and add different kinds of informal awareness cues about the physical world to our online communities. In this paper, we investigate the privacy implications of shared context cues in social networking services. We present an experimental mobile application, which allows users to add different descriptions of context information to their Facebook and Twitter status updates. The application was used by 12 persons during a two-week user trial using their own devices and Facebook accounts. The results indicate that user-defined abstractions of context items were often preferred over more accurate indicators due to privacy concerns or discomfort in sharing. We also found out that using shared context from friends in vicinity needs careful design to overcome the extended privacy implications.
ContextCapture: Exploring the Usage of Context-based Awareness Cues in Inform...Ville Antila
In this paper, we investigate the usage of context-based awareness cues in informal information sharing, especially in social networking services. We present an experimental mobile application, which allows users to add different descriptions of context information to their Facebook status updates. The meaningfulness and the usage of different context descriptions were evaluated in a two-week user trial. The results show that the most frequently used awareness cues in the test setting were location, surroundings, friends and activity. The results also indicate that user-defined semantic abstractions of context items (e.g. “home”, “work”) were often more informative and useful than more accurate indicators (e.g. the address or the name of the place). We also found out that using shared context from friends in vicinity (e.g. identifying the people around) needs careful design to overcome the extended privacy implications.
ContextCapture: Using Context-based Awareness Cues to Create Narrative Events...Ville Antila
In this paper we introduce an experimental application to demonstrate the usage of context-based awareness cues in status updates, especially in SNS’s (Social Networking Services). The presented application allows users to add different descriptions of context information to their Twitter messages and Facebook status updates in a narrative format. We have also developed an adapted version of the system including conference-specific context-types such as the timetable of the presentations and indoor-location detection using Bluetooth beacons. One goal for the demonstrator is to explore the practical use of context abstractions in a conference setup and synthesize interesting insight based on the usage patterns during the event.
A survey on ontology based web personalizationeSAT Journals
Abstract Over the last decade the data on World Wide Web has been growing in an exponential manner. According to Google the data is accelerating with a speed of billion pages per day [24]. Internet has around 2 million users accessing the World Wide Web for various information [25].These numbers certainly raise a severe concern over information over load challenges for the users. Many researchers have been working to overcome the challenge with web personalization, many researchers are looking at ontology based web personalization as an answer to the information overload, as each individual is unique. In this paper we present an overview of ontology based web personalization, Challenges and a survey of the work. This paper also points future work in web personalization. Index Terms: Web Personalization, Ontology, User modeling, web usage mining.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
MindTrek2011 - ContextCapture: Context-based Awareness Cues in Status UpdatesVille Antila
Presentation of an experimental mobile application, which allows users to add different descriptions of context information to their Facebook status updates. The meaningfulness and the usage of different context descriptions were evaluated in a two-week user trial. The results show that the most frequently used awareness cues in the test setting were location, surroundings, friends and activity. The results also indicate that user-defined semantic abstractions of context items (e.g. “home”, “work”) were often more informative and useful than more accurate indicators (e.g. the address or the name of the place). We also found out that using shared context from friends in vicinity (e.g. identifying the people around) needs careful design to overcome the extended privacy implications.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Exploration exploitation trade off in mobile context-aware recommender systemsBouneffouf Djallel
Most existing approaches in Context-Aware Recommender Systems (CRS) focus on recommending relevant items to users taking into account contextual information, such as time, loca-tion, or social aspects. However, none of them have considered the problem of user’s content dynamicity. This problem has been studied in the reinforcement learning community, but without paying much attention to the contextual aspect of the recommendation. We introduce in this paper an algorithm that tackles the user’s content dynamicity by modeling the CRS as a contextual bandit algorithm. It is based on dynamic explora-tion/exploitation and it includes a metric to decide which user’s situation is the most relevant to exploration or exploitation. Within a deliberately designed offline simulation framework, we conduct extensive evaluations with real online event log data. The experimental results and detailed analysis demon-strate that our algorithm outperforms surveyed algorithms.
A contextual bandit algorithm for mobile context-aware recommender systemBouneffouf Djallel
Most existing approaches in Mobile Context-Aware Recommender Systems focus on recommending relevant items to users taking into account contextual information, such as time, location, or social aspects. However, none of them has considered the problem of user’s content evolution. We introduce in this paper an algorithm that tackles this dynamicity. It is based on dynamic exploration/exploitation and can adaptively balance the two aspects by deciding which user’s situation is most relevant for exploration or exploitation. Within a deliberately designed offline simulation framework we conduct evaluations with real online event log data. The experimental results demonstrate that our algorithm outperforms surveyed algorithms.
A survey on context aware system & intelligent Middleware’sIOSR Journals
Abstract: Context aware system or Sentient system is the most profound concept in the ubiquitous computing.
In the cloud system or in distributed computing building a context aware system is difficult task and
programmer should use more generic programming framework. On the basis of layered conceptual design, we
introduce Context aware systems with Context aware middleware’s. On the basis of presented system we will
analyze different approaches of context aware computing. There are many components in the distributed system
and these components should interact with each other because it is the need of many applications. Plenty
Context middleware’s have been made but they are giving partial solutions. In this paper we are giving analysis
of different middleware’s and comprehensive application of it in context caching.
Keywords: Context aware system, Context aware Middleware’s, Context Cache
Agent based frameworks for distributed association rule mining an analysis ijfcstjournal
Distributed Association Rule Mining (DARM) is the task for generating the globally strong association
rules from the global frequent itemsets in a distributed environment. The intelligent agent based model, to
address scalable mining over large scale distributed data, is a popular approach to constructing
Distributed Data Mining (DDM) systems and is characterized by a variety of agents coordinating and
communicating with each other to perform the various tasks of the data mining process. This study
performs the comparative analysis of the existing agent based frameworks for mining the association rules
from the distributed data sources.
An ontology for semantic modelling of virtual worldijaia
This article presents a new representation of semantic virtual environments. We propose to use the ontology as a tool for implementation. Our model, called SVHsIEVs1 provides a consistent representation of the following aspects: the simulated environment, its structure, and the knowledge items using ontology, interactions and tasks that virtual humans can perform in the environment. In SVHsIEVs, we find two type of ontology: the global ontology and the local ontology for Virtual Human. Our architecture has been successfully tested in 3D dynamic environments.
International Journal of Engineering and Science Invention (IJESI) is an international journal intended for professionals and researchers in all fields of computer science and electronics. IJESI publishes research articles and reviews within the whole field Engineering Science and Technology, new teaching methods, assessment, validation and the impact of new technologies and it will continue to provide information on the latest trends and developments in this ever-expanding subject. The publications of papers are selected through double peer reviewed to ensure originality, relevance, and readability. The articles published in our journal can be accessed online.
Understanding interaction context for development interactive modelspaperpublications3
Abstract: We present ongoing research concerning to understand interaction context, its elements and factors emerging during HCI. This paper describes the components during HCI; we analyzed a case study an interactive exhibition in a museum where children are immersed on interaction context. Therefore, in this paper we studied the process of users’ interaction based on user-exhibition interactivity. It gives a general idea, in order to understand how immersed elements can change people’s way interaction negatively or positively.
An Extensible Web Mining Framework for Real KnowledgeIJEACS
With the emergence of Web 2.0 applications that bestow rich user experience and convenience without time and geographical restrictions, web usage logs became a goldmine to researchers across the globe. User behavior analysis in different domains based on web logs has its utility for enterprises to have strategic decision making. Business growth of enterprises depends on customer-centric approaches that need to know the knowledge of customer behavior to succeed. The rationale behind this is that customers have alternatives and there is intense competition. Therefore business community needs business intelligence to have expert decisions besides focusing customer relationship management. Many researchers contributed towards this end. However, the need for a comprehensive framework that caters to the needs of businesses to ascertain real needs of web users. This paper presents a framework named eXtensible Web Usage Mining Framework (XWUMF) for discovering actionable knowledge from web log data. The framework employs a hybrid approach that exploits fuzzy clustering methods and methods for user behavior analysis. Moreover the framework is extensible as it can accommodate new algorithms for fuzzy clustering and user behavior analysis. We proposed an algorithm known as Sequential Web Usage Miner (SWUM) for efficient mining of web usage patterns from different data sets. We built a prototype application to validate our framework. Our empirical results revealed that the framework helps in discovering actionable knowledge.
ContextCapture: Exploring the Usage of Context-based Awareness Cues in Inform...Ville Antila
In this paper, we investigate the usage of context-based awareness cues in informal information sharing, especially in social networking services. We present an experimental mobile application, which allows users to add different descriptions of context information to their Facebook status updates. The meaningfulness and the usage of different context descriptions were evaluated in a two-week user trial. The results show that the most frequently used awareness cues in the test setting were location, surroundings, friends and activity. The results also indicate that user-defined semantic abstractions of context items (e.g. “home”, “work”) were often more informative and useful than more accurate indicators (e.g. the address or the name of the place). We also found out that using shared context from friends in vicinity (e.g. identifying the people around) needs careful design to overcome the extended privacy implications.
ContextCapture: Using Context-based Awareness Cues to Create Narrative Events...Ville Antila
In this paper we introduce an experimental application to demonstrate the usage of context-based awareness cues in status updates, especially in SNS’s (Social Networking Services). The presented application allows users to add different descriptions of context information to their Twitter messages and Facebook status updates in a narrative format. We have also developed an adapted version of the system including conference-specific context-types such as the timetable of the presentations and indoor-location detection using Bluetooth beacons. One goal for the demonstrator is to explore the practical use of context abstractions in a conference setup and synthesize interesting insight based on the usage patterns during the event.
A survey on ontology based web personalizationeSAT Journals
Abstract Over the last decade the data on World Wide Web has been growing in an exponential manner. According to Google the data is accelerating with a speed of billion pages per day [24]. Internet has around 2 million users accessing the World Wide Web for various information [25].These numbers certainly raise a severe concern over information over load challenges for the users. Many researchers have been working to overcome the challenge with web personalization, many researchers are looking at ontology based web personalization as an answer to the information overload, as each individual is unique. In this paper we present an overview of ontology based web personalization, Challenges and a survey of the work. This paper also points future work in web personalization. Index Terms: Web Personalization, Ontology, User modeling, web usage mining.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology.
MindTrek2011 - ContextCapture: Context-based Awareness Cues in Status UpdatesVille Antila
Presentation of an experimental mobile application, which allows users to add different descriptions of context information to their Facebook status updates. The meaningfulness and the usage of different context descriptions were evaluated in a two-week user trial. The results show that the most frequently used awareness cues in the test setting were location, surroundings, friends and activity. The results also indicate that user-defined semantic abstractions of context items (e.g. “home”, “work”) were often more informative and useful than more accurate indicators (e.g. the address or the name of the place). We also found out that using shared context from friends in vicinity (e.g. identifying the people around) needs careful design to overcome the extended privacy implications.
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Exploration exploitation trade off in mobile context-aware recommender systemsBouneffouf Djallel
Most existing approaches in Context-Aware Recommender Systems (CRS) focus on recommending relevant items to users taking into account contextual information, such as time, loca-tion, or social aspects. However, none of them have considered the problem of user’s content dynamicity. This problem has been studied in the reinforcement learning community, but without paying much attention to the contextual aspect of the recommendation. We introduce in this paper an algorithm that tackles the user’s content dynamicity by modeling the CRS as a contextual bandit algorithm. It is based on dynamic explora-tion/exploitation and it includes a metric to decide which user’s situation is the most relevant to exploration or exploitation. Within a deliberately designed offline simulation framework, we conduct extensive evaluations with real online event log data. The experimental results and detailed analysis demon-strate that our algorithm outperforms surveyed algorithms.
A contextual bandit algorithm for mobile context-aware recommender systemBouneffouf Djallel
Most existing approaches in Mobile Context-Aware Recommender Systems focus on recommending relevant items to users taking into account contextual information, such as time, location, or social aspects. However, none of them has considered the problem of user’s content evolution. We introduce in this paper an algorithm that tackles this dynamicity. It is based on dynamic exploration/exploitation and can adaptively balance the two aspects by deciding which user’s situation is most relevant for exploration or exploitation. Within a deliberately designed offline simulation framework we conduct evaluations with real online event log data. The experimental results demonstrate that our algorithm outperforms surveyed algorithms.
A survey on context aware system & intelligent Middleware’sIOSR Journals
Abstract: Context aware system or Sentient system is the most profound concept in the ubiquitous computing.
In the cloud system or in distributed computing building a context aware system is difficult task and
programmer should use more generic programming framework. On the basis of layered conceptual design, we
introduce Context aware systems with Context aware middleware’s. On the basis of presented system we will
analyze different approaches of context aware computing. There are many components in the distributed system
and these components should interact with each other because it is the need of many applications. Plenty
Context middleware’s have been made but they are giving partial solutions. In this paper we are giving analysis
of different middleware’s and comprehensive application of it in context caching.
Keywords: Context aware system, Context aware Middleware’s, Context Cache
Agent based frameworks for distributed association rule mining an analysis ijfcstjournal
Distributed Association Rule Mining (DARM) is the task for generating the globally strong association
rules from the global frequent itemsets in a distributed environment. The intelligent agent based model, to
address scalable mining over large scale distributed data, is a popular approach to constructing
Distributed Data Mining (DDM) systems and is characterized by a variety of agents coordinating and
communicating with each other to perform the various tasks of the data mining process. This study
performs the comparative analysis of the existing agent based frameworks for mining the association rules
from the distributed data sources.
An ontology for semantic modelling of virtual worldijaia
This article presents a new representation of semantic virtual environments. We propose to use the ontology as a tool for implementation. Our model, called SVHsIEVs1 provides a consistent representation of the following aspects: the simulated environment, its structure, and the knowledge items using ontology, interactions and tasks that virtual humans can perform in the environment. In SVHsIEVs, we find two type of ontology: the global ontology and the local ontology for Virtual Human. Our architecture has been successfully tested in 3D dynamic environments.
International Journal of Engineering and Science Invention (IJESI) is an international journal intended for professionals and researchers in all fields of computer science and electronics. IJESI publishes research articles and reviews within the whole field Engineering Science and Technology, new teaching methods, assessment, validation and the impact of new technologies and it will continue to provide information on the latest trends and developments in this ever-expanding subject. The publications of papers are selected through double peer reviewed to ensure originality, relevance, and readability. The articles published in our journal can be accessed online.
Understanding interaction context for development interactive modelspaperpublications3
Abstract: We present ongoing research concerning to understand interaction context, its elements and factors emerging during HCI. This paper describes the components during HCI; we analyzed a case study an interactive exhibition in a museum where children are immersed on interaction context. Therefore, in this paper we studied the process of users’ interaction based on user-exhibition interactivity. It gives a general idea, in order to understand how immersed elements can change people’s way interaction negatively or positively.
An Extensible Web Mining Framework for Real KnowledgeIJEACS
With the emergence of Web 2.0 applications that bestow rich user experience and convenience without time and geographical restrictions, web usage logs became a goldmine to researchers across the globe. User behavior analysis in different domains based on web logs has its utility for enterprises to have strategic decision making. Business growth of enterprises depends on customer-centric approaches that need to know the knowledge of customer behavior to succeed. The rationale behind this is that customers have alternatives and there is intense competition. Therefore business community needs business intelligence to have expert decisions besides focusing customer relationship management. Many researchers contributed towards this end. However, the need for a comprehensive framework that caters to the needs of businesses to ascertain real needs of web users. This paper presents a framework named eXtensible Web Usage Mining Framework (XWUMF) for discovering actionable knowledge from web log data. The framework employs a hybrid approach that exploits fuzzy clustering methods and methods for user behavior analysis. Moreover the framework is extensible as it can accommodate new algorithms for fuzzy clustering and user behavior analysis. We proposed an algorithm known as Sequential Web Usage Miner (SWUM) for efficient mining of web usage patterns from different data sets. We built a prototype application to validate our framework. Our empirical results revealed that the framework helps in discovering actionable knowledge.
Ambiences on the-fly usage of available resources through personal devicesijasuc
In smart spaces such as smart homes, computation is
embedded everywhere: in toys, appliances, or the
home’s infrastructure. Most of these devices provid
e a pool of available resources which the user can
take
advantage, interacting and creating a friendly envi
ronment. The inherent composability of these system
s
and other unique characteristics such as low-cost e
nergy, simplicity in module programming, and even
their small size, make them a suitable candidate fo
r dynamic and adaptive ambient systems. This resear
ch
work focuses on what is defined as an “ambience”, a
space with a user-defined set of computational
devices. A smart-home is modeled as a collection of
ambiences, where every ambience is capable of
providing a pool of available resources to the user
. In turn, the user is supposed to carry one or sev
eral
personal devices able to interact with the ambience
s, taking advantage of his inherent mobility. In th
is way,
the whole system can benefit from resources discove
red in the spatial proximity. A software architectu
re is
designed, which is based on the implementation of l
ow-cost algorithms able to detect and update the sy
stem
when changes in an ambience occur. Ambience middlew
are implementation works in a wide range of
architectures and OSs, while showing a negligible o
verhead in the time to perform the basic output
operations.
Modeling the Adaption Rule in Contextaware Systemsijasuc
Context awareness is increasingly gaining applicability in interactive ubiquitous mobile computing
systems. Each context-aware application has its own set of behaviors to react to context modifications. This
paper is concerned with the context modeling and the development methodology for context-aware systems.
We proposed a rule-based approach and use the adaption tree to model the adaption rule of context-aware
systems. We illustrate this idea in an arithmetic game application.
MODELING THE ADAPTION RULE IN CONTEXTAWARE SYSTEMSijasuc
Context awareness is increasingly gaining applicability in interactive ubiquitous mobile computing
systems. Each context-aware application has its own set of behaviors to react to context modifications. This
paper is concerned with the context modeling and the development methodology for context-aware systems.
We proposed a rule-based approach and use the adaption tree to model the adaption rule of context-aware
systems. We illustrate this idea in an arithmetic game application.
This paper aims to provide an overview of the
contents and design of the all newspapers. Majority of the
newspapers use Blog, RSS and Facebook to connect with
their readers. An online newspaper service providing project.
In this software system users may register as users to read
newspapers online. Once they register they may pay via
dummy credit cards and get access to reading newspapers
online for a month
Recommender systems: a novel approach based on singular value decompositionIJECEIAES
Due to modern information and communication technologies (ICT), it is increasingly easier to exchange data and have new services available through the internet. However, the amount of data and services available increases the difficulty of finding what one needs. In this context, recommender systems represent the most promising solutions to overcome the problem of the so-called information overload, analyzing users' needs and preferences. Recommender systems (RS) are applied in different sectors with the same goal: to help people make choices based on an analysis of their behavior or users' similar characteristics or interests. This work presents a different approach for predicting ratings within the model-based collaborative filtering, which exploits singular value factorization. In particular, rating forecasts were generated through the characteristics related to users and items without the support of available ratings. The proposed method is evaluated through the MovieLens100K dataset performing an accuracy of 0.766 and 0.951 in terms of mean absolute error and root-mean-square error.
A Survey of Building Robust Business Models in Pervasive ComputingOsama M. Khaled
Pervasive computing is one of the most challenging and difficult computing domains nowadays. It includes many architectural challenges like context awareness, adaptability, mobility, availability, and scalability. There are currently few approaches which provide methodologies to build suitable architectural models that are more suited to the nature of the pervasive domain. This area still needs a lot of enhancements in order to let the software business analyst (BA) cognitively handle pervasive applications by using suitable tasks and tools. Accordingly, any proposed research topic that would attempt to define a development methodology can greatly help BAs in modeling pervasive applications with high efficiency. In this survey paper we address some of the most significant and current software engineering practices that are proving to be most effective in building pervasive systems.
For citation:
Osama M. Khaled and Hoda M. Hosny. A Survey of Building Robust Business Models in Pervasive Computing. An accepted paper in the 2014 World Congress in Computer Science Computer Engineering and Applied Computing
Support for Goal Oriented Requirements Engineering in Elastic Cloud Applicationszillesubhan
Businesses have already started to exploit potential uses of cloud computing as a new paradigm for promoting their services. Although the general concepts they practically focus on are: viability, survivability, adaptability, etc., however, on the ground, there is still a lack for forming mechanisms to sustain viability with adaptation of new requirements in cloud-based applications. This has inspired a pressing need to adopt new methodologies and abstract models which support system acquisition for self-adaptation, thus guaranteeing autonomic cloud application behavior. This paper relies over state-of-the-art Neptune framework as runtime adaptive software development environment supported with intention-oriented modeling language in the representation and adaptation of goal based model artifacts and their intrinsic properties requirements. Such an approach will in turn support distributed service based applications virtually over the cloud to sustain a self-adaptive behavior with respect to its functional and non-functional characteristics.
Activity Context Modeling in Context-AwareEditor IJCATR
The explosion of mobile devices has fuelled the advancement of pervasive computing to provide personal assistance in this
information-driven world. Pervasive computing takes advantage of context-aware computing to track, use and adapt to contextual
information. The context that has attracted the attention of many researchers is the activity context. There are six major techniques that
are used to model activity context. These techniques are key-value, logic-based, ontology-based, object-oriented, mark-up schemes and
graphical. This paper analyses these techniques in detail by describing how each technique is implemented while reviewing their pros
and cons. The paper ends with a hybrid modeling method that fits heterogeneous environment while considering the entire of modeling
through data acquisition and utilization stages. The modeling stages of activity context are data sensation, data abstraction and
reasoning and planning. The work revealed that mark-up schemes and object-oriented are best applicable at the data sensation stage.
Key-value and object-oriented techniques fairly support data abstraction stage whereas the logic-based and ontology-based techniques
are the ideal techniques for reasoning and planning stage. In a distributed system, mark-up schemes are very useful in data
communication over a network and graphical technique should be used when saving context data into database.
A HUMAN-CENTRIC APPROACH TO GROUP-BASED CONTEXT-AWARENESSIJNSA Journal
The emerging need for qualitative approaches in context-aware information processing calls for proper modelling of context information and efficient handling of its inherent uncertainty resulted from human interpretation and usage. Many of the current approaches to context-awareness either lack a solid theoretical basis for modelling or ignore important requirements such as modularity, high-order uncertainty management and group-based context-awareness. Therefore, their real-world application and extendibility remains limited. In this paper, we present f-Context as a service-based contextawareness framework, based on language-action perspective (LAP) theory for modelling. Then we identify some of the complex, informational parts of context which contain high-order uncertainties due to differences between members of the group in defining them. An agent-based perceptual computer architecture is proposed for implementing f-Context that uses computing with words (CWW) for handling uncertainty. The feasibility of f-Context is analyzed using a realistic scenario involving a group of mobile users. We believe that the proposed approach can open the door to future research on context-awareness by offering a theoretical foundation based on human communication, and a service-based layered architecture which exploits CWW for context-aware, group-based and platform-independent access to information systems.
A USER PROFILE BASED ACCESS CONTROL MODEL AND ARCHITECTUREIJCNC
Personalization and adaptation to the user profile capability are the hottest issues to ensure ambient
assisted living and context awareness in nowadays environments. With the growing healthcare and
wellbeing context aware applications, modeling security policies becomes an important issue in the
design of future access control models. This requires rich semantics using ontology modeling for the
management of services provided to dependant people. However, current access control models remain
unsuitable due to lack of personalization, adaptability and smartness to the handicap situation.
Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...Ramesh Iyer
In today's fast-changing business world, Companies that adapt and embrace new ideas often need help to keep up with the competition. However, fostering a culture of innovation takes much work. It takes vision, leadership and willingness to take risks in the right proportion. Sachin Dev Duggal, co-founder of Builder.ai, has perfected the art of this balance, creating a company culture where creativity and growth are nurtured at each stage.
Epistemic Interaction - tuning interfaces to provide information for AI supportAlan Dix
Paper presented at SYNERGY workshop at AVI 2024, Genoa, Italy. 3rd June 2024
https://alandix.com/academic/papers/synergy2024-epistemic/
As machine learning integrates deeper into human-computer interactions, the concept of epistemic interaction emerges, aiming to refine these interactions to enhance system adaptability. This approach encourages minor, intentional adjustments in user behaviour to enrich the data available for system learning. This paper introduces epistemic interaction within the context of human-system communication, illustrating how deliberate interaction design can improve system understanding and adaptation. Through concrete examples, we demonstrate the potential of epistemic interaction to significantly advance human-computer interaction by leveraging intuitive human communication strategies to inform system design and functionality, offering a novel pathway for enriching user-system engagements.
GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...James Anderson
Effective Application Security in Software Delivery lifecycle using Deployment Firewall and DBOM
The modern software delivery process (or the CI/CD process) includes many tools, distributed teams, open-source code, and cloud platforms. Constant focus on speed to release software to market, along with the traditional slow and manual security checks has caused gaps in continuous security as an important piece in the software supply chain. Today organizations feel more susceptible to external and internal cyber threats due to the vast attack surface in their applications supply chain and the lack of end-to-end governance and risk management.
The software team must secure its software delivery process to avoid vulnerability and security breaches. This needs to be achieved with existing tool chains and without extensive rework of the delivery processes. This talk will present strategies and techniques for providing visibility into the true risk of the existing vulnerabilities, preventing the introduction of security issues in the software, resolving vulnerabilities in production environments quickly, and capturing the deployment bill of materials (DBOM).
Speakers:
Bob Boule
Robert Boule is a technology enthusiast with PASSION for technology and making things work along with a knack for helping others understand how things work. He comes with around 20 years of solution engineering experience in application security, software continuous delivery, and SaaS platforms. He is known for his dynamic presentations in CI/CD and application security integrated in software delivery lifecycle.
Gopinath Rebala
Gopinath Rebala is the CTO of OpsMx, where he has overall responsibility for the machine learning and data processing architectures for Secure Software Delivery. Gopi also has a strong connection with our customers, leading design and architecture for strategic implementations. Gopi is a frequent speaker and well-known leader in continuous delivery and integrating security into software delivery.
Essentials of Automations: Optimizing FME Workflows with ParametersSafe Software
Are you looking to streamline your workflows and boost your projects’ efficiency? Do you find yourself searching for ways to add flexibility and control over your FME workflows? If so, you’re in the right place.
Join us for an insightful dive into the world of FME parameters, a critical element in optimizing workflow efficiency. This webinar marks the beginning of our three-part “Essentials of Automation” series. This first webinar is designed to equip you with the knowledge and skills to utilize parameters effectively: enhancing the flexibility, maintainability, and user control of your FME projects.
Here’s what you’ll gain:
- Essentials of FME Parameters: Understand the pivotal role of parameters, including Reader/Writer, Transformer, User, and FME Flow categories. Discover how they are the key to unlocking automation and optimization within your workflows.
- Practical Applications in FME Form: Delve into key user parameter types including choice, connections, and file URLs. Allow users to control how a workflow runs, making your workflows more reusable. Learn to import values and deliver the best user experience for your workflows while enhancing accuracy.
- Optimization Strategies in FME Flow: Explore the creation and strategic deployment of parameters in FME Flow, including the use of deployment and geometry parameters, to maximize workflow efficiency.
- Pro Tips for Success: Gain insights on parameterizing connections and leveraging new features like Conditional Visibility for clarity and simplicity.
We’ll wrap up with a glimpse into future webinars, followed by a Q&A session to address your specific questions surrounding this topic.
Don’t miss this opportunity to elevate your FME expertise and drive your projects to new heights of efficiency.
Key Trends Shaping the Future of Infrastructure.pdfCheryl Hung
Keynote at DIGIT West Expo, Glasgow on 29 May 2024.
Cheryl Hung, ochery.com
Sr Director, Infrastructure Ecosystem, Arm.
The key trends across hardware, cloud and open-source; exploring how these areas are likely to mature and develop over the short and long-term, and then considering how organisations can position themselves to adapt and thrive.
Encryption in Microsoft 365 - ExpertsLive Netherlands 2024Albert Hoitingh
In this session I delve into the encryption technology used in Microsoft 365 and Microsoft Purview. Including the concepts of Customer Key and Double Key Encryption.
Dev Dives: Train smarter, not harder – active learning and UiPath LLMs for do...UiPathCommunity
💥 Speed, accuracy, and scaling – discover the superpowers of GenAI in action with UiPath Document Understanding and Communications Mining™:
See how to accelerate model training and optimize model performance with active learning
Learn about the latest enhancements to out-of-the-box document processing – with little to no training required
Get an exclusive demo of the new family of UiPath LLMs – GenAI models specialized for processing different types of documents and messages
This is a hands-on session specifically designed for automation developers and AI enthusiasts seeking to enhance their knowledge in leveraging the latest intelligent document processing capabilities offered by UiPath.
Speakers:
👨🏫 Andras Palfi, Senior Product Manager, UiPath
👩🏫 Lenka Dulovicova, Product Program Manager, UiPath
UiPath Test Automation using UiPath Test Suite series, part 3DianaGray10
Welcome to UiPath Test Automation using UiPath Test Suite series part 3. In this session, we will cover desktop automation along with UI automation.
Topics covered:
UI automation Introduction,
UI automation Sample
Desktop automation flow
Pradeep Chinnala, Senior Consultant Automation Developer @WonderBotz and UiPath MVP
Deepak Rai, Automation Practice Lead, Boundaryless Group and UiPath MVP
The Art of the Pitch: WordPress Relationships and SalesLaura Byrne
Clients don’t know what they don’t know. What web solutions are right for them? How does WordPress come into the picture? How do you make sure you understand scope and timeline? What do you do if sometime changes?
All these questions and more will be explored as we talk about matching clients’ needs with what your agency offers without pulling teeth or pulling your hair out. Practical tips, and strategies for successful relationship building that leads to closing the deal.
The Art of the Pitch: WordPress Relationships and Sales
Following the user’s interests in mobile context aware recommender systems
1. Following the User’s Interests in Mobile Context-Aware
Recommender Systems : The hybrid-ε-greedy algorithm
Djallel Bouneffouf, Amel Bouzeghoub & Alda Lopes Gançarski
Department of Computer Science, Télécom SudParis, UMR CNRS Samovar, 91011 Evry Cedex, France
{Djallel.Bouneffouf, Amel.Bouzeghoub, Alda.Gancarski}@it-sudparis.eu
Abstract— The wide development of mobile applications based on his current context. This goal matches with the goal
provides a considerable amount of data of all types (images, of recommender systems.
texts, sounds, videos, etc.). In this sense, Mobile Context-aware Both context-aware systems and recommender systems
Recommender Systems (MCRS) suggest the user suitable are used to provide users with relevant information and/or
information depending on her/his situation and interests. Two services; the first based on the user’s context; the second
key questions have to be considered 1) how to recommend the based on the user’s interests. Therefore, the next logical step
user information that follows his/her interests evolution? 2) is to combine these two systems within the so-called mobile
how to model the user’s situation and its related interests? To context-aware recommender systems.
the best of our knowledge, no existing work proposing a MCRS
The notion of context is not new. In general, it refers to
tries to answer both questions as we do. This paper describes
an ongoing work on the implementation of a MCRS based on
any information that is relevant to the user and his/her
the hybrid-ε-greedy algorithm we propose, which combines the surroundings. The main difficulty when dealing with context
standard ε-greedy algorithm and both content-based filtering is its high dynamicity and its constant change. A change in
and case-based reasoning techniques. one contextual feature may generate changes in other
contextual features. That’s why context-aware systems have
Keywords-component; context-aware recommender systems; to be adapted to highly dynamic user situations and context
personalization; situation. models need to capture the dynamic interactions between
contextual information.
I. INTRODUCTION Several works in context-aware systems have addressed
the problem of context modeling and interesting single or
Mobile technologies have made access to a huge hybrid models exist. Most of them are gathered in the survey
collection of information, anywhere and anytime. Thereby, done in [5]. However, these models have not been widely
information is customized according to users’ needs and applied to the mobile application domain. The dynamic
preferences. This brings big challenges for the interactions between contextual dimensions and user profiles
Recommender System field. Indeed, technical features of have not been fully represented by the existing models which
mobile devices yield to navigation practices which are more often focus on modeling static associations among different
difficult than the traditional navigation task. contextual information.
Recommender systems are systems that produce In recommender systems, a considerable amount of
individualized suggestions concerning interesting content the research has been done in recommending relevant
user might like out of a large number of alternatives. Often, information for mobile users. Earlier techniques ([12], [15])
recommender systems using collaborative or content-based are based solely on the computational behavior of the user to
filtering algorithms are applied. model his interests regardless of his surrounding
In mobile applications, information personalization is environment (location, time, nearby people). The main
even more important, because of the limitations of mobile limitation of such approaches is that they do not take into
devices in terms of displays, input capabilities, bandwidth account the contextual evolution of the user interests w. r. t.
etc. It is then desirable to personalize not only using pre- his context.
defined user profiles, but also the user’s context such as the This study gives rise to another category of
current location. recommendation techniques that tackle these limitations
We can find similar challenges in context-aware systems when building situation-aware user profiles. Two key
area. A general definition of context-aware systems is given questions have to be considered, namely 1) how to
in [4]: “A system is context-aware if it uses context to recommend the user information that follows his/her
provide relevant information and/or services to the user, interests evolution? 2) how to model the user’s situation and
where relevancy depends on the user’s task.” its related user’s interests?
The key goal of context-aware systems is hence to
provide a user with relevant information and/or services
2. In order to tackle these problems, our approach consists documents may be chosen. However, obtaining information
of: about the documents’ average rewards (i.e., exploration) can
Capturing context through communication with refine B’s estimate of the documents’ rewards and in turn
diverse information sources: from the mobile device increase long-term user’s satisfaction. Clearly, neither a
GPS capability, if it is possible, or by inferring a purely exploring nor a purely exploiting algorithm works
user’s location based on the user’s calendar or other best in general, and a good tradeoff is needed. The authors
devices in the neighborhood. on [7, 8, 16] describe a smart way to balance exploration and
Storing history for future use or user habit analysis. exploitation. However, none of them consider the user’s
Reasoning and learning capabilities to be able to (i) situation during the recommendation.
use the collected contextual information for mapping B. Modeling the situation and user profile
low-level information to symbolic contextual
information (e.g. mapping an exact coordinate to a Few research works are dedicated to manage the user’s
symbolic location such as an address) or to deduce situation on recommendation. In [2, 9, 10] the authors
higher level information i.e. identify user situation propose a method which consists of building a dynamic
and recommend the most relevant content according user’s profile based on time and user’s experience. The
to the situation (e.g. determine that the user is in a user’s preferences in the user’s profile are weighted
room where a meeting is in progress so the cell according to the situation (time, location) and the user’s
phone should be switched to silent mode); (ii) learn behavior. To model the evolution on the user’s preferences
the behavior based on reasoning and the history data according to his temporal situation in different periods, (like
(e.g. whenever the user enters a room which has a workday or vacations), the weighted association for the
meeting in progress or where the next event is a concepts in the user’s profile is established for every new
meeting, the user turns the volume of the cell phone experience of the user. The user’s activity combined with the
down and searches for the meeting agenda. The user's profile are used together to filter and recommend
system may learn this behavior for the next time). relevant content.
Another work [6] describes a MCRS operating on three
The remainder of the paper is organized as follows. dimensions of context that complement each other to get
Section II reviews some related works. Section III presents highly targeted. First, the MCRS analyzes information such
the user model. Section IV introduces the recommendation as clients’ address books to estimate the level of social
algorithm and the last section concludes the paper and points affinity among users. Second, it combines social affinity with
out possible directions for future work. the spatiotemporal dimensions and the user’s history in order
to improve the quality of the recommendations.
II. BACKGROUND In [1], the authors present a technique to perform user-
based collaborative filtering. Each user’s mobile device
We mention now recent relevant recommendation
stores all explicit ratings made by its owner as well as ratings
techniques that tackle the two issues mentioned above,
received from other users. Only users in spatiotemporal
namely: following the evolution of user’s interests and
proximity are able to exchange ratings and they show how
modeling the situation and user profile.
this provides a natural filtering based on social contexts.
A. Following the evolution of user’s interests Each work cited above tries to recommend interesting
The trend today on recommender systems is to suggest information to users on contextual situation; however they
relevant information to users, using supervised machine do not consider the evolution of the user interest.
learning techniques. In these approaches, the recommender
system has to execute two steps: (1) the learning step, where To summarize, none of the mentioned works tackles both
the system learns from samples and gradually adjusts its problems. As a result, our approach exploits the following
parameters; (2) the exploitation step, where new samples are new features:
presented to the system to perform a generalization [18]. We consider user situations as a multi-dimensional
These approaches suffer from the following drawbacks: space where each dimension is represented by a
(i) need for initial information about the user’s interests domain ontology. Unlike in [1, 2, 6, 10], where
provided by an expert; (ii) difficulty in following the context items are low level data, in our approach
evolution of the user’s interests. Some works found in the they correspond to concepts of social, location and
literature [7, 8, 16] address these problems as a need for time ontologies. Each situation is associated to
balancing exploration and exploitation studied in the “bandit specific user profile and preferences.
algorithm”. A bandit algorithm B exploits its past experience Inspired by models of human reasoning developed
to select documents that appear more frequently. Besides, by [14] in robotic, we propose to consider user's
these seemingly optimal documents may in fact be situation in the bandit algorithm by using case-based
suboptimal, due to imprecision in B’s knowledge. In order to reasoning technique, which is not considered in [7,
avoid this undesired situation, B has to explore documents 8, 9, 16].
by actually choosing seemingly suboptimal documents so as In [7, 8, 16] authors use a smart bandit algorithm to
to gather more information about them. Exploitation can manage the exploration/exploitation strategy,
increase short-term user’s satisfaction since some suboptimal however they do not take into account the content in
3. the strategy. Our intuition is that, considering the - The indirect preference: it is the information that we
content when managing the exploration/exploitation extract from the user system interaction, for example the
strategy will improve it. This is why we propose to number of clicks on the visited documents or the time spent
use content-based filtering techniques together with on a document.
ε-greedy algorithm. Let UP be the preferences submitted by a specific user in
the system at a given situation. Each document in UP is
In what follows, we define the structure of the proposed represented as a single vector d=(c1,...,cn), where ci (i=1, ..,
user model and the methods for inferring the n) is the value of a component characterizing the preferences
recommendation situations. Then, we explain how to build of d. We consider the following components: the total
dynamic user profiles, and how to manage the number of clicks on d, the total time spent reading d, the
exploration/exploitation strategy, according to the current number of times d was recommended, and the direct
situation. preference rate on d.
2) History
All the interactions between the user and the system are
III. THE PROPOSED USER’S MODEL stored as well as the situations in order to exploit this data to
As stated in Section II, static approaches for building the improve the recommendation process.
user’s profile [12, 15] are poorly useful in our context, so we calendar Before each meeting, the user has to fill-up the
rather focus on more dynamic techniques, capable of calendar with information concerning the person(s) to meet,
continuously adjusting the user’s interests to the current Time and Location instances are automatically inferred by
situation. the system.
Figure 1 depicts the diagram of the proposed user model.
B. User Situation
A situation S can be represented as a triple whose
features X are the values assigned to each dimension: S = (Xl,
Xt, Xs), where Xl (resp. Xt and Xs) is the value of the location
(resp. time and social) dimension.
Suppose the user is associated to: the location
"38.868143, 2.3484122" from his phone’s GPS; the time
"Mon Oct 3 12:10:00 2011" from his phone’s watch; and the
meeting with Paul Gerard from his calendar. To build the
situation, we associate to this kind of low level data, directly
acquired from mobile devices capabilities, more abstracted
concepts using ontologies reasoning means.
1) Location
There are different ways to characterize a location. As
Figure 1. User model diagram
returned by location sensor systems (like GPS), location is a
A. User Profile position in systems based on geographic coordinates, or may
also be defined by an address. Simple automated place
The user profile is composed of the user’s personal data and labeling systems are already commercialized (Google maps,
other dynamic information, including his preferences, his Yahoo local...) and consist of merging data such as postal
calendar and the history of his interactions with the system. addresses with maps.
1) UserPreferences In our user model, we use a local spatial ontology to
The user preferences are contextual and might depend on represent and reason on geographic information. Using this
many factors, like the location or the current task within an ontology, for the above example, we get, from location
activity. Thus, they are associated to the user situation and "38.86, 2.34", the value “Paris” to insert in the location
the user activity. Preferences are deduced during user dimension of the situation.
navigation activities. They contain the set of navigation 2) Time
documents used by the user in a situation. A navigation The temporal information is complex: it is continuous
activity expresses the following sequence of events: (i) the and can be represented at different levels of granularity. To
user logs in the system and navigates across documents to define the temporal aspects characterizing the user’s
get the desired information; (ii) the user expresses his situation, we suggest to abstract the continuum time into
preferences on the visited documents. We assume that a specific and significant periods (abstract time classes), which
visited document is relevant, and thus belongs to the user’s we expect having an effect on the user’s behavior (e.g.
preferences, if there are some observable user’s behaviors morning, weekend). To allow a good representation of the
through 2 types of preference: temporal information and its manipulation, we propose to
- The direct preference: the user expresses his interest in use OWL-Time ontology [11] which is today a reference for
the document by inserting a rate, like for example putting representing and reasoning about time. We propose to base
stars (“*”) at the top of the document. our work on this ontology and extend it if necessary. Taking
the example above, for the time value "Mon Oct 3 12:10:00
4. 2011", we get, using the OWL-Time ontology, the value defines the exploration/exploitation tradeoff; ε is the
“workday”. probability of recommending a random exploratory
3) Social connection document.
The social connection refers to the information of the
user’s interlocutors (e. g. a friend, an important customer, a Algorithm 1 The ε-greedy algorithm
colleague or his manager). To define the neighborhood 1: Input: ε, UPc, N
people aspects characterizing the user, a clear model for the 2: Output: D
representation and reasoning on social clustering is
necessary. We use the FOAF Ontology [13] to describe the 3: D = Ø
social network by a set of concepts and properties. For 4: For i =1 to N do
example, the information about “the meeting with Paul 5: q = Random({0,1})
Gerard” can yield the value “wine client” for the social
dimension.
6: arg max (getCTR(d)) if q ≤ ε
d(UP D )
IV. THE PROPOSED RECOMMENDATION ALGORITHM 7: di =
In our MCRS, documents’ recommendation is modeled 8: Random(UPc) otherwise
as a multi-armed bandit problem. Formally, a bandit 9: D= D ∪ di
algorithm proceeds in discrete rounds t = 1,… T. For each 10: Endfor
round t, the algorithm performs the following tasks:
11: Return D
Task 1. It observes the current user’s situation St and a set
Dt of documents with their feature vectors xt,d for d Dt. The
To select the document to be recommended, there are
vector xt,d gives information of both user’s situation St and
several strategies which provide an approximate solution for
document d.
the bandit problem. Here, we focus on two of them (as in
Task 2. Based on observed rewards in previous rounds, it
[16]):
chooses a document dt Dt, and receives reward rt ,d whose
t - the greedy strategy (Alg. 1, line 6), which estimates each
expectation depends on both the user’s situation St and the document’s CTR; then it always chooses the best documents,
document dt. i.e. the ones having the higher CTR; thus it only performs
Task 3. It then improves its document-selection strategy exploitation;
with the new observation. - the ε-greedy strategy, which adds some greedy exploration
In tasks 1 to 3, the total T-round reward of D is defined policy to the greedy strategy, choosing the best document
as T r with probability 1–ε (Alg. 1, line 6) or a randomly picked
t 1 t , d
t
document otherwise (probability ε) (Alg. 1, line 8).
B. The proposed hybrid-ε-greedy algorithm
while the optimal expected T-round reward is defined as
t 1 rt ,d *
T
We propose a two-fold improvement on the performance
t
of the ε-greedy algorithm: integrating case base reasoning
where dt* is the document with maximum expected reward at (CBR) and content based filtering (CBF). This new proposed
round t. Our goal is to design the bandit algorithm so that the algorithm is called hybrid-ε-greedy and is described in
expected total reward is maximized. Algorithm 4.
In the field of document recommendation, when a To improve exploitation of the ε-greedy algorithm, we
document is presented to the user and this one selects it by a propose to integrate CBR into each iteration: before choosing
click, a reward of 1 is incurred; otherwise, the reward is 0. the document, the algorithm computes the similarity between
With this definition of reward, the expected reward of a the present situation and each one in the situation base; if
document is precisely its Click Through Rate (CTR). The there is a situation that can be re-used, the algorithm
CTR is the average number of clicks on a recommended retrieves it, and then applies an exploration/exploitation
document, computed diving the total number of clicks on it strategy.
by the number of times it was recommended. In this situation-aware computing approach, the premise
A. The ε-greedy algorithm part of a case is a specific situation S of a mobile user when
he navigates on his mobile device, while the value part of a
The ε-greedy strategy is sketched in Algorithm 1. For a
case is the user’s preferences UP to be used for the
given user’s situation, the algorithm recommends a pre-
recommendation. Each case from the case base is denoted as
defined number of documents, specified by the parameter N.
C= (S, UP).
In this algorithm, UPc={d1,…,dP} is the set of documents
Let Sc=(Xlc, Xtc, Xsc) be the current situation of the user,
corresponding to the current user’s preferences;
UP the current user’s preferences and PS={S1,....,Sn} the set
c
D={d1,….,dN} is the set of documents to recommend;
of past situations. The proposed hybrid-ε-greedy algorithm
getCTR (Alg. 1, line 6) is the function which estimates the
involves the following four methods.
CTR of a given document; Random (Alg. 1, lines 5 and 8) is
the function returning a random element from a given set; q 1) RetriveCase() (Alg. 4, line 4)
is a random value uniformly distributed over [0, 1] which
5. Given the current situation Sc, the RetrieveCase method
determines the expected user preferences by comparing Sc
with the situations in past cases in order to choose the most Algorithm 2 The RecommendDocuments() method
similar one Ss. The method returns, then, the corresponding 1: Input: ε, UPc, N
case (Ss, UPs). 2: Output: D
Ss is selected from PS by computing the following 3: D=Ø
expression as it done by [9]: 4: For i=1 to N do
S s = arg max α j sim j X c ,X ij (1) 5: q = Random({0, 1})
j
S PS
i
j 6: j = Random({0, 1})
In equation 1, simj is the similarity metric related to 7: arg max (getCTR(d)) if j<q<ε
dimension j between two situation vectors and αj the weight d(UP D )
associated to dimension j. αj is not considered in the scope of 8: di = CBF(UPc-D, arg max (getCTR(d)) if q≤ j≤ε
this paper, taking a value of 1 for all dimensions. d(UP D )
The similarity between two concepts of a dimension j in 9: Random(UPc) otherwise
an ontological semantic depends on how closely they are 10: D = D ∪ {di }
related in the corresponding ontology (location, time or 11: Endfor
social). We use the same similarity measure as [17] defined 12: Return D
by equation 2:
sim X c , X i 2
deph( LCS ) (2) Algorithm 3 The CBF() algorithm
(deph( X c ) deph( X ij ))
j j j
j 1: Input: UP, db
Here, LCS is the Least Common Subsumer of Xjc and Xji, and 2: Output: ds
depth is the number of nodes in the path from the node to the 3: ds= arg max (cossim( db , d))
ontology root. dUP
2) RecommendDocuments() (Alg. 4, line 6) 4: Return ds
In order to insure a better precision of the recommender
results, the recommendation takes place only if the following
condition is verified: sim(Sc, Ss) ≥ B (Alg. 4, line 5), where B 3) UpdateCase() & InsertCase()
is a threshold value and After recommending documents applying the
sim(S c , S s ) = sim j X c ,X s
j j
RecommendDocuments method (Alg. 4, line 6), the user’s
j preferences are updated w. r. t. number of clicks and number
In the RecommendDocuments() method, sketched in of recommendations for each recommended document on
Algorithm 2, we propose to improve the ε-greedy strategy by which the user clicked at least one time. This is done by the
applying CBF in order to have the possibility to recommend, UpdatePreferences function (Alg. 4, line 7).
not the best document, but the most similar to it (Alg. 2, line Depending on the similarity between the current situation
8). We believe this may improve the user’s satisfaction. Sc and its most similar situation Ss (computed with
The CBF algorithm (Algorithm 3) computes the RetriveCase()), being 3 the number of dimensions in the
similarity between each document d=(c1,..,ck) from UP context, two scenarios are possible:
(except already recommended documents D) and the best - sim(Sc, Ss) ≠ 3: the current situation does not exist in the
document db=(cjb ,.., ckb ) and returns the most similar one. case base (Alg. 4, line 8); the InsertCase() method adds to
The degree of similarity between d and db is determined by the case base the new case composed of the current situation
using the cosine measure, as indicated in equation 3: Sc and the updated UP.
ck ck (3) - sim(Sc, Ss) = 3: the situation exists in the case base (Alg. 4,
d db line 10); the UpdateCase() method updates the case having
cos sim(d , d b ) k
d db premise situation Sc with the updated UP.
c
k
b
k
2
c kb2
k
6. Algorithm 4 hybrid-ε-greedy algorithm
1. Input: B, ε, N, PS, Ss, UPs, Sc, UPc
2. Output: D
3. D = Ø [4] Dey, A. K. Providing Architectural Support for Building Context-
Aware Applications. Ph.D. thesis, College of Computing, Georgia
4. (Ss , UPs) = RetriveCase(Sc, PS) Institute of Technology, 2000.
5. if sim(Sc,Ss) ≥ B then [5] Dobson, S. Leveraging the subtleties of location, in Proc. of Smart
Objects and Ambient Intelligence, 2005.
6. D = RecommendDocuments(ε, UPs, N)
[6] Lakshmish R, Deepak P, Ramana P, Kutila G, Dinesh Garg, Karthik
7. UPc = UpdatePreferences(UPs, D) V, Shivkumar K. Tenth International Conference on Mobile Data
Management: Systems, Services and Middleware CAESAR: A
8. if sim(Sc, Ss) ≠ 3 then Mobile context-aware, Social Recommender System for Low-End
9. PS = InsertCase(Sc, UPc) Mobile Devices, 2009.
10. else [7] Lihong Li, Presented at the Nineteenth International Conference on
World Wide Web, Raleigh, NC, USA, 2010 A Contextual-Bandit
11. PS = UpdateCase(Sp, UPc) Approach to Personalized News Article Recommendation, 2010.
12. end if [8] Mieczysław A. Kłopotek. Approaches to Cold-Start in recommender
systems. studia informatica Systems and information technology,
13. else PS = InsertCase(Sc, UPc); 2009.
14. end if [9] Ourdia, B.,Tamine, L., Boughanem M. Dynamically Personalizing
Search Results for Mobile Users: FQAS'09, pp. 99-110. In
15. Return D Proceedings of the 8th International Conference on Flexible Query
Answering Systems (2009).
V. CONCLUSION [10] Panayiotou, C., Samaras, G. Mobile User Personalization with
This paper describes our approach for implementing a Dynamic Profiles. On the Move to Meaningful Internet Systems,
MCRS. Our contribution is to make a deal between 2006.
exploration and exploitation for learning and maintaining [11] Pan, F. Representing complex temporal phenomena for the semantic
web and natural language. Ph.D thesis, University of Southern
user’s interests based on his navigation history. California, Dec, 2007.
In the future, we plan to evaluate our approach on real [12] Samaras, G., Panayiotou, C. Personalized portals for the wireless user
mobile navigation contexts obtained from a diary study based on mobile agents. Proc. 2nd Int'l workshop on Mobile
conducted with collaboration with the Nomalys French Commerce, pp. 70-74, 2002.
company. This study will yield to know if considering the [13] Sparck-Jones, K. Index Term Weighting, Information Storage and
situation on the exploration/exploitation strategy Retrieval, vol. 9, no. 2011, pp. 619-633, 1973.
significantly increases the performance of the system on [14] Taylor E and PeterStone. Transfer learning for bandit algorithm
following the user’s interests. domains: A survey. Journal of Machine Learning Research,
10(1):1633–1685, 2009.
[15] Varma, V., Sriharsha, N., Pingali, P. Personalized web search engine
REFERENCES for mobile devices, Int'l Workshop on Intelligent Information Access,
IIIA'06, 2006.
[1] Alexandre S, Moira C. Norrie, Michael Grossniklaus and Beat Signer, [16] Wei Li ,Exploitation and Exploration in a Performance based
“Spatio-Temporal Proximity as a Basis for Collaborative Filtering in Contextual Advertising SystemProceedings of the 16th ACM
Mobile Environments”.Workshop on Ubiquitous Mobile Information SIGKDD international conference on Knowledge discovery and data
and Collaboration Systems, CAiSE, 2006. miningACM New York, NY, USA, 2010.
[2] Bellotti, V., Begole, B., Chi, E.H., Ducheneaut, N., Fang, J., Isaacs, [17] Wu, Z., Palmer, M. Verb Semantics and Lexical Selection. In
E. Activity-Based Serendipitous Recommendations with the Magitti Proceedings of the 32nd Annual Meeting of the Association for
Mobile Leisure Guide. Proceedings On the Move, 2008. Computational Linguistics, 1994.
[3] Bettini, C., et al., A survey of context modelling and reasoning [18] Zhang T. and Iyengar, V. “Recommender systems using linear
techniques, Pervasive and Mobile Computing, 2009.
classifiers,” The Journal of Machine Learning Research, vol. 2, p.
334, 2002.