Towards a Context Definition for Multi-Agent Systems

This chapter aims to define context notion for multi-agent systems (MAS). Starting from the state of the art on context in different disciplines, we present context as a generic and abstract notion. We argue that context depends on three characteristics: domain, entity, and problem. By specifying this definition with MAS, we initially consider context from an extensional point of view as three components—actant, role, and situation—and then from an intensional one, which represents the context model for agents in MAS which consist of information on environment, other objects, agents, and relations between them. Therefore, we underline a new way of representing agent knowledge, building context on this knowledge, and using it. Furthermore, we prove the applicability of contextual agent solution for other research fields, particularly in personalized information retrieval by taking into account as agents: crawlers and as objects: documents.

Role-Based Multi-Agent Systems

Agent system design is a complex task challenging designers to simulate intelligent collaborative behavior. Roles can reduce the complexity of agent system design by categorizing the roles played by agents. The role concepts can also be used in agent systems to describe the collaboration among cooperative agents. In this chapter, we introduce roles as a means to support interaction and collaboration among agents in multi-agent systems. We review the application of roles in current agent systems at first, then describe the fundamental principles of role-based collaboration and propose the basic methodologies of how to apply roles into agent systems (i.e., the revised E-CARGO model). After that, we demonstrate a case study: a soccer robot team designed with role specifications. Finally, we present the potentiality to apply roles into information personalization.

Privacy Control Requirements for Context-Aware Mobile Services

With the rapid developments of mobile telecommunications technology over the last two decades, a new computing paradigm known as ‘anywhere and anytime’ or ‘ubiquitous’ computing has evolved. Consequently, attention has been given not only to extending current Web services and mobile service models and architectures, but increasingly also to make these services context-aware. Privacy represents one of the hot topics that has questioned the success of these services. In this chapter, we discuss the different requirements of privacy control in context-aware services architectures. Further, we present the different functionalities needed to facilitate this control. The main objective of this control is to help end users make consent decisions regarding their private information collection under conditions of uncertainty. The proposed functionalities have been prototyped and integrated in a UMTS locationbased mobile services testbed platform on a university campus. Users have experienced the services in real time. A survey of users’ responses on the privacy functionality has been carried out and analyzed as well. Users’ collected response on the privacy functionality was positive in most cases. Additionally, results obtained reflected the feasibility and usability of this approach.

Clustering Web Information Sources

The explosive growth of the Web scale has drastically increased information circulation and dissemination rates. As the number of both Web users and Web sources grows significantly everyday, crucial data management issues, such as clustering on the Web, should be addressed and analyzed. Clustering has been proposed towards improving both the information availability and the Web users’ personalization. Clusters on the Web are either users’ sessions or Web information sources, which are managed in a variation of applications and implementations testbeds. This chapter focuses on the topic of clustering information over the Web, in an effort to overview and survey on the theoretical background and the adopted practices of most popular emerging and challenging clustering research efforts. An up-to-date survey of the existing clustering schemes is given, to be of use for both researchers and practitioners interested in the area of Web data mining.

Mining for Web Personalization

The Web has become a huge repository of information and keeps growing exponentially under no editorial control, while the human capability to find, read and understand content remains constant. Providing people with access to information is not the problem; the problem is that people with varying needs and preferences navigate through large Web structures, missing the goal of their inquiry. Web personalization is one of the most promising approaches for alleviating this information overload, providing tailored Web experiences. This chapter explores the different faces of personalization, traces back its roots and follows its progress. It describes the modules typically comprising a personalization process, demonstrates its close relation to Web mining, depicts the technical issues that arise, recommends solutions when possible, and discusses the effectiveness of personalization and the related concerns. Moreover, the chapter illustrates current trends in the field suggesting directions that may lead to new scientific results.

On Personalizing Web Services Using Context

This chapter presents a context-based approach for Web services personalization so that user preferences are accommodated. Preferences are of different types, varying from when the execution of a Web service should start to where the outcome of this execution should be delivered according to user location. Besides user preferences, it will be discussed in this chapter that the computing resources on which the Web services operate have an impact on their personalization. Indeed, resources schedule the execution requests that originate from multiple Web services. To track the personalization of a Web service from a temporal perspective (i.e., what did happen, what is happening, and what will happen), three types of contexts are devised and referred to as user context, Web service context, and resource context.

Service-Oriented Architectures for Context-Aware Information Retrieval and Access

Humans are quite successful at conveying ideas to each other and retrieving information from interactions appropriately. This is due to many factors: the richness of the language they share, the common understanding of how the world works, and an implicit understanding of everyday situations (Dey & Abowd, 1999). When humans talk with humans, they are able to use implicit situational information (i.e., context) to enhance the information exchange process. Context (Cool & Spink, 2002) plays a vital part in adaptive and personalized information retrieval and access. Unfortunately, computer communications lacks this ability to provide auxiliary context in addition to the substantial content of information. As computers are becoming more and more ubiquitous and mobile, there is a need and possibility to provide information “personalized, any time, and anywhere” (ITU, 2006). In these scenarios, large amounts of information circulate in order to create smart and proactive environments that will significantly enhance both the work and leisure experiences of people. Context-awareness plays an important role to enable personalized information retrieval and access according to the current situation with minimal human intervention. Although context-aware information retrieval systems have been researched for a decade (Korkea-aho, 2000), the rise of mobile and ubiquitous computing put new challenges to issue, and therefore we are motivated to come up with new solutions to achieve non-intrusive, personalized information access on the mobile service platforms and heterogeneous wireless environments.

Personalized Content-Based Image Retrieval

In content-based image retrieval (CBIR), a set of low-level features are extracted from an image to represent its visual content. Retrieval is performed by image example where a query image is given as input by the user and an appropriate similarity measure is used to find the best matches in the corresponding feature space. This approach suffers from the fact that there is a large discrepancy between the low-level visual features that one can extract from an image and the semantic interpretation of the image’s content that a particular user may have in a given situation. That is, users seek semantic similarity, but we can only provide similarity based on low-level visual features extracted from the raw pixel data, a situation known as the semantic gap. The selection of an appropriate similarity measure is thus an important problem. Since visual content can be represented by different attributes, the combination and importance of each set of features varies according to the user’s semantic intent. Thus, the retrieval strategy should be adaptive so that it can accommodate the preferences of different users. Relevance feedback (RF) learning has been proposed as a technique aimed at reducing the semantic gap. It works by gathering semantic information from user interaction. Based on the user’s feedback on the retrieval results, the retrieval scheme is adjusted. By providing an image similarity measure under human perception, RF learning can be seen as a form of supervised learning that finds relations between high-level semantic interpretations and low-level visual properties. That is, the feedback obtained within a single query session is used to personalize the retrieval strategy and thus enhance retrieval performance. In this chapter we present an overview of CBIR and related work on RF learning. We also present our own previous work on a RF learning-based probabilistic region relevance learning algorithm for automatically estimating the importance of each region in an image based on the user’s semantic intent.

User and Context-Aware Quality Filters Based on Web Metadata Retrieval

Due to the amount of information on the Web being so large and being of varying levels of quality, it is becoming increasingly difficult to find precisely what is required on the Web, particularly if the information consumer does not have precise knowledge of his or her information needs. On the Web, while searching for information, users can find data that is old, imprecise, invalid, intentionally wrong, or biased, due to this large amount of available data and comparative ease of access. In this environment users constantly receive useless, outdated, or false data, which they have no means to assess. This chapter addresses the issues regarding the large amount and low quality of Web information by proposing a methodology that adopts user and context-aware quality filters based on Web metadata retrieval. This starts with an initial evaluation and adjusts it to consider context characteristics and user perspectives to obtain aggregated evaluation values.

Overview of Design Options for Neighborhood-Based Collaborative Filtering Systems

The problem of collaborative filtering is to predict how well a user will like an item that he or she has not rated, given a set of historical ratings for this and other items from a community of users. A plethora of collaborative filtering algorithms have been proposed in related literature. One of the most prevalent families of collaborative filtering algorithms are neighborhood-based ones, which calculate a prediction of how much a user will like a particular item, based on how other users with similar preferences have rated this item. This chapter aims to provide an overview of various proposed design options for neighborhood-based collaborative filtering systems, in order to facilitate their better understanding, as well as their study and implementation by recommender systems’ researchers and developers. For this purpose, the chapter extends a series of design stages of neighborhood-based algorithms, as they have been initially identified by related literature on collaborative filtering systems. Then, it reviews proposed alternatives for each design stage and provides an overview of potential design options.