Exploiting rich user information for one-class collaborative filtering

One-class collaborative filtering (OCCF) is an emerging setup in collaborative filtering in which only positive examples or implicit feedback can be observed. Compared with the traditional collaborative filtering setting where the data have ratings, OCCF is more realistic in many scenarios when no ratings are available. In this paper, we propose to improve OCCF accuracy by exploiting the rich user information that is often naturally available in community-based interactive information systems, including a user’s search query history, and purchasing and browsing activities. We propose two major strategies to incorporate such user information into the OCCF models: One is to linearly combine scores from different sources, and the other is to embed user information into collaborative filtering. Furthermore, we employ the MapReduce framework for similarity computation over millions of users and items. Experimental results on two large-scale retail datasets from a major e-commerce company show that the proposed methods are effective and can improve the performance of the OCCF over baseline methods through leveraging rich user information.     

基于关键词相关性的有害信息爬虫系统研究

传统的互联网有害信息发现方法是依据Google、百度等元搜索工具,用户输入关键词进行检索,然后对获取的结果进行研判,但是用户经常无法准确地描述所需的资料,给出的关键词不准确,搜索结果常有用户不关心的垃圾数据,而一些有用的数据却不能列出。文中探讨了一种基于元搜索,引入关键词扩充技术的爬虫方法。该方法在网页抓取,用户检索的时候能扩充输入的关键词,从而提高搜索覆盖率和精度。该方法投入小,效果好,还可通过扩展应用到其他领域。     

Feature-based and Clique-based User Models for Movie Selection: A Comparative Study

The huge amount of information available in the currently evolving world wide information infrastructure at any one time can easily overwhelm end-users. One way to address the information explosion is to use an ‘information filtering agent’ which can select information according to the interest and/or need of an end-user. However, at present few information filtering agents exist for the evolving world wide multimedia information infrastructure. In this study, we evaluate the use of feature-based approaches to user modeling with the purpose of creating a filtering agent for the video-on-demand application. We evaluate several feature and clique-based models for 10 voluntary subjects who provided ratings for the movies. Our preliminary results suggest that feature-based selection can be a useful tool to recommend movies according to the taste of the user and can be as effective as a movie rating expert. We compare our feature-based approach with a clique-based approach, which has advantages where information from other users is available. This revised version was published online in July 2006 with corrections to the Cover Date.     

Ontology based user query interpretation for semantic multimedia contents retrieval

Users who are familiar with the existing keyword-based search have problems of not being able to configure the formal query because they don’t have generic knowledge on knowledge base when using the semantic-based retrieval system. User wants the search results which are more accurate and match the user’s search intents with the existing keyword-based search and the same search keyword without the need to recognize what technology the currently used retrieval system is based on to provide the search results. In order to do the semantic analysis of the ambiguous search keyword entered by users who are familiar with the existing keyword-based search, ontological knowledge base constructed based on refined meta-data is necessary, and the keyword semantic analysis technique which reflects user’s search intents from the well-established knowledge base and can generate accurate search results is necessary. In this paper, therefore, by limiting the knowledge base construction to multimedia contents meta-data, the applicable prototype has been implemented and its performance in the same environment as Smart TV has been evaluated. Semantic analysis of user’s search keyword is done, evaluated and recommended through the proposed ontological knowledge base framework so that accurate search results that match user’s search intents can be provided.     

基于用户动机模型的搜索引擎研究

提出一种基于用户动机模型的网络搜索引擎和一种提高用户行为模型构建效率的方案.动机模型建立于用户与搜索引擎之间,用以辅助用户检索,以达到提高搜索引擎检索效率和准确率的目的.以人类行为学为理论基础,以个性化技术为手段,从而合并相似的用户行为模型以构建用户动机模型.通过实验,验证了基于用户动机模型的搜索引擎比通用搜索引擎能更好地适应用户的需求.     

Adaptation and user expertise modelling in AthosMail

This article describes the User Model component of AthosMail, a speech-based interactive e-mail application developed in the context of the EU project DUMAS. The focus is on the system’s adaptive capabilities and user expertise modelling, exemplified through the User Model parameters dealing with initiative and explicitness of the system responses. The purpose of the conducted research was to investigate how the users could interact with a system in a more natural way, and the two aspects that mainly influence the system’s interaction capabilities, and thus the naturalness of the dialogue as a whole, are considered to be the dialogue control and the amount of information provided to the user. The User Model produces recommendations of the system’s appropriate reaction depending on the user’s observed competence level, monitored and computed on the basis of the user’s interaction with the system. The article also discusses methods for the evaluation of adaptive user models and presents results from the AthosMail evaluation.The research was done while the author was affiliated with the University of Art and Design Helsinki as the scientific coordinator of the DUMAS project.     

A collaborative filtering framework based on both local user similarity and global user similarity

Collaborative filtering as a classical method of information retrieval has been widely used in helping people to deal with information overload. In this paper, we introduce the concept of local user similarity and global user similarity, based on surprisal-based vector similarity and the application of the concept of maximin distance in graph theory. Surprisal-based vector similarity expresses the relationship between any two users based on the quantities of information (called surprisal) contained in their ratings. Global user similarity defines two users being similar if they can be connected through their locally similar neighbors. Based on both of Local User Similarity and Global User Similarity, we develop a collaborative filtering framework called LS&GS. An empirical study using the MovieLens dataset shows that our proposed framework outperforms other state-of-the-art collaborative filtering algorithms.     

Mobile mail-agents through similarity-based reasoning

 Bots, or software agents are programs designed to perform tasks autonomously. Mailbots attempt to provide useful functions about electronic mail (E-mail) service such as filtering information, gathering information, and scheduling. With Internet use continuing to explode, the information overload is growing so fast that the same virtues that made E-mail so popular are now becoming a negative technologic “boomerang” (see the volume of junk or spam mail). Industrial as well as academic research has faced this problem in terms of automated filtering methods in order to distinguish legtimate E-mail from spamming. Here we describe an alternative approach: our mailbot is skilled to find “appropriate” destination of the message triggering a spidering process on an Intranet-based network. The spidering performs a distributed, mobile computation via pervasive agents: by applying a similarity-based reasoning on designed users resources the agents are able to deduct if the contacted user may be interested or not in receiving the E-mail. The overall architecture is implemented in Java using the basic issues of Internet protocol.     

A conceptual model for user-centered quality information retrieval on the World Wide Web

Information retrieval from the Internet is becoming a commonplace phenomenon. Users and consumers are browsing websites and seeking various kinds of information for personal use. Retrieving quality information from the Internet can be challenging even for the computer-savvy. There are several search engines, even some personalized, to help users search for information on the Internet. In spite of all the claims about search engines, users still have difficult time retrieving relevant information quickly. This paper proposes a general conceptual model for user-centered quality information retrieval (UCQIR) from the Internet. The UCQIR conceptual model is presented in an architectural form. The UCQIR architectural model uses the concept of “Task-performer” to present various aspects of an information retrieval system at the knowledge level. Task-performer is an abstract construct used to conceptualize the idea of an entity that is competent in doing its tasks. The UCQIR architectural model can be used to easily design and develop domain-specific, user-centered quality information retrieval systems. The proposed UCQIR conceptual model is unique and comprehensive. The use of the conceptual model is illustrated through a design of a patient-centered quality medical information retrieval for the medical domain. We also present an experimental evaluation of a UCQIR prototype based upon real user experiences. The experimental results are very positive.     

Integration of document index with perception index

 We propose the perception index (PI) that contains attributes associated with a focal keyword restricted by fuzzy term(s) used in fuzzy queries on the Internet. The PI assists the user to reflect his/her perception in the process of query. If we integrate the document index (DI) used in commercial Web search engines with the proposed PI, we can handle both crisp terms (keyword-based) and fuzzy terms (perception-based). In this respect, the proposed approach is softer than the keyword-based approach. The PI brings somewhat closer to natural language. It is a further step toward a real human-friendly, natural language-based interface for Internet. It should greatly help the user relatively easily retrieve relevant information.     

Interest based selection of user generated content for rich communication services

The last few years, we have witnessed an exponential growth in available content, much of which is user generated (e.g. pictures, videos, blogs, reviews, etc.). The downside of this overwhelming amount of content is that it becomes increasingly difficult for users to identify the content they really need, resulting into considerable research efforts concerning personalized search and content retrieval.On the other hand, this enormous amount of content raises new possibilities: existing services can be enriched using this content, provided that the content items used match the user's personal interests. Ideally, these interests should be obtained in an automatic, transparent way for an optimal user experience.In this paper two models representing user profiles are presented, both based on keywords and with the goal to enrich real-time communication services. The first model consists of a light-weight keyword tree which is very fast, while the second approach is based on a keyword ontology containing extra temporal relationships to capture more details of the user's behavior, however exhibiting lower performance. The profile models are supplemented with a set of algorithms, allowing to learn user interests and retrieving content from personal content repositories.In order to evaluate the performance, an enhanced instant messaging communication service was designed. Through simulations the two models are assessed in terms of real-time behavior and extensibility. User evaluations allow to estimate the added value of the approach taken. The experiments conducted indicate that the algorithms succeed in retrieving content matching the user's interests and both models exhibit a linear scaling behavior. The algorithms perform clearly better in finding content matching several user interests when benefiting from the extra temporal information in the ontology based model.     

中文搜索引擎的智能探讨

目前各中文网站上的搜索引擎都是基于关键词的，其对用户输入信息与资源信息库是精确匹配，这是造成中文搜索引擎低智能的主要原因。分析了关键词搜索引擎的利弊，根据人们输入信息的不确定性，其在网上检索信息的过程应该是智能化的特点，提出了提高中文搜索引擎智能的方法。     

基于用户本体模型的文本信息过滤方法研究

随着互联网信息的快速剧增,文本过滤技术成为互联网内容处理的关键技术,对海量信息处理具有很重要的意义.目前研究热点是基于语义的过滤方法,但是这些方法一般都需要大量规则和领域知识的支持,可用性不是很好.为了使机器更好地理解用户需求和文本内容,使过滤结果更能反映用户的真正需求,提高文本过滤的准确率和召回率,提出了基于用户本体模型UOM的文本信息过滤方法.该方法主要包括UOM构建、文本结构分析、文本概念提取和语义相关度计算等.基于UOM(User Ontology Model)的过滤方法,不仅可以表示复杂的用户需求,而且还避免了领域本体的构建,因而其有效性和实用性得到了很大的提高.通过在网络教学资源的智能按需服务系统中的实际运用,表明此方法能更有效地为用户提供过滤服务.     

Rational behavior in peer-to-peer profile obfuscation for anonymous keyword search

User profiling in web search has the advantage of enabling personalized web search: the quality of the results offered by the search engine to the user is increased by taking the user’s interests into account when presenting those results. The negative side is that the interests and the query history of users may contain information considered as private; hence, technology should be provided for users to avoid profiling if they wish so. There are several anti-profiling approaches in web search, from basic level countermeasures to private information retrieval and including profile obfuscation. Except private information retrieval (PIR), which hides the retrieved item from the database, the rest of approaches focus on anonymizing the user’s identity and fall into the category of anonymous keyword search (also named sometimes user-private information retrieval). Most current PIR protocols are ill-suited to provide PIR from a search engine or large database, due to their complexity and their assumption that the database actively cooperates in the PIR protocol. Peer-to-peer profile obfuscation protocols appear as a competitive option provided that peers are rationally interested in helping each other. We present a game-theoretic analysis of P2P profile obfuscation protocols which shows under which conditions helping each other is in the peers’ rational interest.     

Modeling user interests by conceptual clustering

As more information becomes available on the Web, there has been a crescent interest in effective personalization techniques. Personal agents providing assistance based on the content of Web documents and the user interests emerged as a viable alternative to this problem. Provided that these agents rely on having knowledge about users contained into user profiles, i.e., models of user preferences and interests gathered by observation of user behavior, the capacity of acquiring and modeling user interest categories has become a critical component in personal agent design. User profiles have to summarize categories corresponding to diverse user information interests at different levels of abstraction in order to allow agents to decide on the relevance of new pieces of information. In accomplishing this goal, document clustering offers the advantage that an a priori knowledge of categories is not needed, therefore the categorization is completely unsupervised. In this paper we present a document clustering algorithm, named WebDCC (Web Document Conceptual Clustering), that carries out incremental, unsupervised concept learning over Web documents in order to acquire user profiles. Unlike most user profiling approaches, this algorithm offers comprehensible clustering solutions that can be easily interpreted and explored by both users and other agents. By extracting semantics from Web pages, this algorithm also produces intermediate results that can be finally integrated in a machine-understandable format such as an ontology. Empirical results of using this algorithm in the context of an intelligent Web search agent proved it can reach high levels of accuracy in suggesting Web pages.     

The virtual library secretary: A user model-based software agent

A user model comprises knowledge of the user's past and present tasks, and is the essential element in adaptive user interfaces. Through the propagation of user models, we can take the user models into the world of software agents, and thus construct user model-based software agents. The user model approach reduces the threats of software agents penetrating a local host and the amount of data transferred. This paper presents the Virtual Library Secretary, which is a user model-based software agent system. The Virtual Library Secretary offers information retrieval and information filtering to the user. The system is part of the Virtual Secretary project. The user model is established by a simple neural network. In this way, the agent is able to learn and adapt to the user's behaviour. This paper discusses the user model concept, presents the Virtual Secretary system architecture and describes how this architecture works through the Virtual Library Secretary.The Virtual Secretary project is an on-going project at the University of Tromsø. It includes two phases: the first phase (ViSe) focuses on user model-based software agents for information filtering and agent control propagation; the second phase (ViSe2) concentrates on information integration via cooperative agents in a distributed environment. The project is partly supported by the Research Council of Norway (Grant no. 112577/431).     

基于知识的网页检索工具

随着因特网在全球范围的广泛使用,越来越多的人们借助于因特网从事科研和商务活动,而网页检索工具成了人们必不可少的软件工具．然而,目前流行的检索工具大多基于关键字查询,常常出现信息过载或有用信息丢失等现象．造成这一原因主要有两方面：用户提交的查询不能很好地表达他的目的;查询的结果没有建立有效的索引机制,引导人们快速找到有用信息。为此我们提出一种基于知识的网页检索工具（ＫＷＳＥ）,它是在已有的检索工具的     

信息过滤中用户兴趣模型的表示

用户兴趣模型的表示是信息过滤中的关键技术,它直接关系到过滤效果的好坏。比较了三种不同的用户兴趣模型:传统的关键字表示法、固定文章集法F D S (文章表示法)和基于示例法(段落表示法) ,并对它们进行了实验分析,得出后两种方法要优于第一种的结论。其技术的关键在于如何描述和更新用户的兴趣模型。     

Anatomy and Empirical Evaluation of an Adaptive Web-Based Information Filtering System

A case study in adaptive information filtering systems for the Web is presented. The described system comprises two main modules, named HUMOS and WIFS. HUMOS is a user modeling system based on stereotypes. It builds and maintains long term models of individual Internet users, representing their information needs. The user model is structured as a frame containing informative words, enhanced with semantic networks. The proposed machine learning approach for the user modeling process is based on the use of an artificial neural network for stereotype assignments. WIFS is a content-based information filtering module, capable of selecting html/text documents on computer science collected from the Web according to the interests of the user. It has been created for the very purpose of the structure of the user model utilized by HUMOS. Currently, this system acts as an adaptive interface to the Web search engine ALTA VISTA^TM. An empirical evaluation of the system has been made in experimental settings. The experiments focused on the evaluation, by means of a non-parametric statistics approach, of the added value in terms of system performance given by the user modeling component; it also focused on the evaluation of the usability and user acceptance of the system. The results of the experiments are satisfactory and support the choice of a user model-based approach to information filtering on the Web.     

Buyer agent to enhance consumer awareness: SAATHI

Personal agents have been developed that assist a user with information processing needs by generating, filtering, collecting, or transforming information. On the other hand, internet stores are providing services customized by the needs and interests of individual customers. Such services can be viewed as “seller’s agents” whose goal is to push merchandise and/or services on to the users. This leads us to believe that there is a growing need for deploying “buyer’s agents” whose goal is to best serve the consumer’s interests. The Internet contains a huge volume of information which can overwhelm a buyer. The buyers may often make misinformed decisions based on partial, outdated, irrelevant or incorrect information. We have identified several key functionalities of buyer’s agents whose goal is to reduce information overload and improve relevancy and accuracy of information for consumers. In particular, such agents can make consumers aware of complex interactions between specified preferences and prevailing market conditions, provide differential analysis for decision support, and use domain ontologies to help the user reformulate queries to improve satisfaction with query results. We present SAATHI, a prototype buyer’s agent that demonstrate some of these functionalities in an apartment locator domain.