Personalization: Learning User Profiles (original) (raw)
Related papers
Personalization in Digital Libraries: An Intelligent Service based on Semantic User Profiles
Suppose you registered to a large scientific congress and you got from the Web site the conference program containing a long list of papers which will be presented. Which presentations do you choose to attend? Usually either you try to guess the most interesting talks from their titles and authors or you are forced to have a quick look at the conference proceedings. A recommender system able to learn your research interests from the latest papers you wrote or read, and use them to provide suggestions, might be of valuable help for you in this scenario. Content-based recommenders analyze documents previously rated by a target user, and build a profile exploited to recommend new interesting documents. One of the main limitations of traditional keyword-based approaches is that they are unable to capture the semantics of the user interests, due to the natural language ambiguity. We developed a semantic recommender system, called ITem Recommender 1 , able to disambiguate documents before using them to learn the user profile. The Conference Participant Advisor service relies on the profiles learned by ITem Recommender to build a personalized conference program, in which relevant talks are highlighted according to the participant interests.
User Profiles for Personalized Information Access
Lecture Notes in Computer Science, 2007
The amount of information available online is increasing exponentially. While this information is a valuable resource, its sheer volume limits its value. Many research projects and companies are exploring the use of personalized applications that manage this deluge by tailoring the information presented to individual users. These applications all need to gather, and exploit, some information about individuals in order to be effective. This area is broadly called user profiling. This chapter surveys some of the most popular techniques for collecting information about users, representing, and building user profiles. In particular, explicit information techniques are contrasted with implicitly collected user information using browser caches, proxy servers, browser agents, desktop agents, and search logs. We discuss in detail user profiles represented as weighted keywords, semantic networks, and weighted concepts. We review how each of these profiles is constructed and give examples of projects that employ each of these techniques. Finally, a brief discussion of the importance of privacy protection in profiling is presented. This chapter discusses user profiles specifically designed for providing personalized information access. Other types of profiles, build using different construction techniques, are described elsewhere in this book. In particular, Chapter 4 [40] discusses generic user modeling systems that are broader in scope, not necessarily focused on Internet applications. Related research on collaborative recommender systems, discussed in Chapter 9 of this book [81], combines information from multiple users in order to provide improved information services. Concern over privacy protection is growing in parallel with the demand for personalized features. These two trends seem to be in direct opposition to each other, so privacy protection must be a crucial component of every personalization system. A detailed discussion can be found in Chapter 21 of this book [39]. There are a wide variety of applications to which personalization can be applied and a wide variety of different devices available on which to deliver the personalized information. Early personalization research focused on personalized filtering and/or rating systems for e-mail [49], electronic newspapers [14, 16], Usenet newsgroups [41, 58, 86, 91, 106], and Web documents [4]. More recently, personalization efforts have focused on improving navigation effectiveness by providing browsing assistants [9, 13], and adaptive Web sites [69]. Because search is one of the most common activities performed today, many projects are now focusing on personalized Web search [46, 88, 92] and more details on the subject can be found in Chapter 6 of this book [52]. However, personalized approaches to searching other types of collections, e.g., short stories [76], Java source code [100], and images [14] have also been explored. Commercial products are also adopting personalized features, for example, Yahoo!'s personalized Web portals [110] and Google Lab's personalized search [30]. The aforementioned systems are just a few examples that illustrate the breadth of applications to which personalized approaches are being investigated. Nichols [63] and Oard and Marchionini [64] provide a general overview of some the issues and approaches to personalized rating and filtering and Pretschner [71] describes approximately 45 personalization systems. Most personalization systems are based on some type of user profile, a data instance of a user model that is applied to adaptive interactive systems. User profiles may include demographic information, e.g., name, age, country, education level, etc, and may also represent the interests or preferences of either a group of users or a single person. Personalization of Web portals, for example, may focus on individual users, for example, displaying news about specifically chosen topics or the market summary of specifically selected stocks, or a groups of users for whom distinctive characteristics where identified, for example, displaying targeted advertising on ecommerce sites. In order to construct an individual user's profile, information may be collected explicitly, through direct user intervention, or implicitly, through agents that monitor user activity. Although profiles are typically built only from topics of interest to the user, some projects have explored including information about non-relevant topics in the profile [35, 104]. In these approaches, the system is able to use both kinds of topics to identify relevant documents and discard non-relevant documents at the same time. Profiles that can be modified or augmented are considered dynamic, in contrast to static profiles that maintain the same information over time. Dynamic profiles that Explicit info Data Collection Technology Or Application Profile Constructor User Implicit info Keyword profile Semantic Net profile Concept profile Personalized Services As shown in Figure 2.1, the user profiling process generally consists of three main phases. First, an information collection process is used to gather raw information about the user. As described in Section 2.2, depending on the information collection process selected, different types of user data can be extracted. The second phase focuses on user profile construction from the user data. Section 2.3 summarizes a variety of ways in which profiles may be represented and Section 2.4 some of the ways a profile may be constructed. The final phase, in which a technology or application exploits information in the user profile in order to provide personalized services, is discussed in Parts II and III of this book. 2.2 Collecting Information About Users The first phase of a profiling technique collects information about individual users. A basic requirement of such a system is that it must be able to uniquely identify users. This task is described in more detail in Section 2.2.1. The information collected may be explicitly input by the user or implicitly gathered by a software agent. It may be collected on the user's client machine or gathered by the application server itself. Depending on how the information is collected, different data about the users may be extracted. Several options, and their impacts, are discussed in Section 2.2.2. In
Personalization in Digital Library: An Intelligent Service based on Semantic User Profiles
Abstract—Suppose you registered to a large scientific congress and you got from the Web site the conference program containing a long list of papers which will be presented. Which presentations do you choose to attend? Usually either you try to guess the most interesting talks from their titles and authors or you are forced to have a quick look at the conference proceedings.
Web Information Personalization: Challenges and Approaches
Lecture Notes in Computer Science, 2003
As the number of web pages increases dramatically, the problem of the information overload becomes more severe when browsing and searching the WWW. To alleviate this problem, personalization becomes a popular remedy to customize the Web environment towards a user's preference. To date, recommendation systems and personalized web search systems are the most successful examples of Web personalization. By focusing on these two types of systems, this paper reviews the challenges and the corresponding approaches proposed in the past ten years.
A Review on Personalization Techniques
International Journal of Recent Research Aspects, 2015
The World Wide Web is commonly known as "web" is a huge collection of interlinked billions of HTML documents. The automated searching tools such as search Engines are used to retrieve information from such a huge collection on web. Although the present search engines are using sophisticated ranking and indexing algorithms, but they still provide a long list of documents, most of which are not relevant to the user's need. One of the reason behind this is each user has different requirement for information search. So, it becomes vital in many areas to consider the user interests and preferences for the search engine while retrieving and maintaining its database. Personalization is taken as one of the method to provide relevant information according to the user's need. This paper conducts a survey of how personalization can provide useful knowledge to the user. Several user personalization approaches and techniques are illustrated in this paper.
Dynamic user profiles for web personalisation
Expert Systems with Applications, 2015
Web personalization systems are used to enhance the user experience by providing tailor-made services based on the user's interests and preferences which are typically stored in user profiles. For such systems to remain effective, the profiles need to be able to adapt and reflect the users' changing behaviour. In this paper, we introduce a set of methods designed to capture and track user interests and maintain dynamic user profiles within a personalization system. User interests are represented as ontological concepts which are constructed by mapping web pages visited by a user to a reference ontology and are subsequently used to learn short-term and long-term interests. A multi-agent system facilitates and coordinates the capture, storage, management and adaptation of user interests. We propose a search system that utilizes our dynamic user profile to provide a personalized search experience. We present a series of experiments that show how our system can effectively model a dynamic user profile and is capable of learning and adapting to different user browsing behaviours.
Automated user modeling for personalized digital libraries
International Journal of Information Management, 2006
Digital libraries (DL) have become one of the most typical ways of accessing any kind of digitalized information. Due to this key role, users welcome any improvements on the services they receive from digital libraries. One trend used to improve digital services is through personalization. Up to now, the most common approach for personalization in digital libraries has been user-driven. Nevertheless, the design of efficient personalized services has to be done, at least in part, in an automatic way. In this context, machine learning techniques automate the process of constructing user models. This paper proposes a new approach to construct digital libraries that satisfy user's necessity for information: Adaptive Digital Libraries, libraries that automatically learn user preferences and goals and personalize their interaction using this information.
2010
Recently, personalised search engines and recommendation systems have been widely adopted by users who require assistance in searching, classifying, and filtering information. This paper presents an overview of the field of personalisation systems and describes current state-of-the-art methods and techniques. It reviews approaches for (1) user profiling, including behaviour, preference, and intention modelling; (2) content modelling, comprising content representation, analysis, and classification; and (3) filtering methods for recommendation, classified into four main categories: rule-based, contentbased, collaborative, and hybrid filtering. The paper also discusses personalisation systems in different domains, and various techniques and their limitations. Finally, it identifies several issues and possible directions for further research that can improve recommendation capabilities and enhance personalised systems.
2003
The World Wide Web has become one of the prime mechanisms for information dissemination and has experienced tremendous growth resulting in an information explosion. This information overload is exaggerated by the quantity of information that is available as well as the quality and format of that information. This paper highlights the positive impact that personalization may have in alleviating information overload and the additional benefits it holds for e-business.
Adaptive User Modeling for Personalization of Web Contents
2004
This paper presents a system for personalization of web contents based on a user model that stores long term and short term interests. Long term interests are modeled through the selection of specific and general categories, and keywords for which the user needs information. However, user needs change over time as a result of his interaction with received information. For this reason, the user model must be capable of adapting to those shifts in interest. In our case, this adaptation of the user model is performed by a short term model obtained from user provided feedback. The evaluation performed with 100 users during 15 days has determined that the combined use of long and short term models performs best when specific and general categories and keywords are used together for the long term model.