This paper presents an approach to automatically optimizing the retrieval quality of search engines using clickthrough data. Intuitively, a good information retrieval system should present relevant documents high in the ranking, with less relevant documents following below. While previous approaches to learning retrieval functions from examples exist, they typically require training data generated from relevance judgments by experts. This makes them difficult and expensive to apply. The goal of this paper is to develop a method that utilizes clickthrough data for training, namely the query-log of the search engine in connection with the log of links the users clicked on in the presented ranking. Such clickthrough data is available in abundance and can be recorded at very low cost. Taking a Support Vector Machine (SVM) approach, this paper presents a method for learning retrieval functions. From a theoretical perspective, this method is shown to be well-founded in a risk minimization framework. Furthermore, it is shown to be feasible even for large sets of queries and features. The theoretical results are verified in a controlled experiment. It shows that the method can effectively adapt the retrieval function of a meta-search engine to a particular group of users, outperforming Google in terms of retrieval quality after only a couple of hundred training examples.

The World Wide Web (WWW) continues to grow at an astounding rate in both the sheer volume of tra#c and the size and complexity of Web sites. The complexity of tasks such as Web site design, Web server design, and of simply navigating through a Web site have increased along with this growth. An important input to these design tasks is the analysis of how a Web site is being used. Usage analysis includes straightforward statistics, such as page access frequency, as well as more sophisticated forms of analysis, such as finding the common traversal paths through a Web site. Web Usage Mining is the application of data mining techniques to usage logs of large Web data repositories in order to produce results that can be used in the design tasks mentioned above. However, there are several preprocessing tasks that must be performed prior to applying data mining algorithms to the data collected from server logs. This paper presents several data preparation techniques in order to identify unique users and user sessions. Also, a method to divide user sessions into semantically meaningful transactions is defined and successfully tested against two other methods. Transactions identified by the proposed methods are used to discover association rules from real world data using the WEBMINER system [15].

The World Wide Web is a vast source of information accessible to computers, but understandable only to humans. The goal of the research described here is to automatically create a computer understandable world wide knowledge base whose content mirrors that of the World Wide Web. Such a

Abstract not found

The World Wide Web is a vast source of information accessible to computers, but understandable only to humans. The goal of the research described here is to automatically create a computer understandable knowledge base whose content mirrors that of the World Wide Web. Such a knowledge base would enable much more effective retrieval of Web information, and promote new uses of the Web to support knowledge-based inference and problem solving. Our approach is to develop a trainable information extraction system that takes two inputs. The first is an ontology that defines the classes (e.g., company, person, employee, product) and relations (e.g., employed_by, produced_by) of interest when creating the knowledge base. The second is a set of training data consisting of labeled regions of hypertext that represent instances of these classes and relations. Given these inputs, the system learns to extract information from other pages and hyperlinks on the Web. This article describes our general a...

The World-Wide Web is developing very fast. Currently, finding useful information on the Web is a time consuming process. In this paper, we present WebMate, an agent that helps users to effectively browse and search the Web. WebMate extends the state of the art in Web-based information retrieval in many ways. First, it uses multiple TF-IDF vectors to keep track of user interests in different domains. These domains are automatically learned by WebMate. Second, WebMate uses the Trigger Pair Model to automatically extract keywords for refining document search. Third, during search, the user can provide multiple pages as similarity/relevance guidance for the search. The system extracts and combines relevant keywords from these relevant pages and uses them for keyword refinement. Using these techniques, WebMate provides effective browsing and searching help and also compiles and sends to users personal newspaper by automatically spiding news sources. We have experimentally evaluated the per...

... In this paper we consider the use of implicit feedback techniques for query expansion and user profiling in information retrieval tasks. These techniques unobtrusively obtain information about users by watching their natural interactions with the system. Some of the user behaviors that have been most extensively investigated as sources of implicit feedback include reading time, saving, printing and selecting. The primary advantage to using implicit techniques is that such techniques remove the cost to the user of providing feedback. Implicit measures are generally thought to be less accurate than explicit measures [Nic97], but as large quantities of implicit data can be gathered at no extra cost to the user, they are attractive alternatives. Moreover, implicit measures can be combined with explicit ratings to obtain a more accurate representation of user interests. Implicit

Abstract not found

Domain-specific internet portals are growing in popularity because they gather content from the Web and organize it for easy access, retrieval and search. For example, www.campsearch.com allows complex queries by age, location, cost and specialty over summer camps. This functionality is not possible with general, Web-wide search engines. Unfortunately these portals are difficult and time-consuming to maintain. This paper advocates the use of machine learning techniques to greatly automate the creation and maintenance of domain-specific Internet portals. We describe new research in reinforcement learning, information extraction and text classification that enables efficient spidering, the identification of informative text segments, and the population of topic hierarchies. Using these techniques, we have built a demonstration system: a portal for computer science research papers. It already contains over 50,000 papers and is publicly available at www.cora.justresearch.com. These techniques are ...

Inspired by Hill and Hollan's original work [6], we have been developing a theory of interaction history and building tools to apply this theory to navigation in a complex information space. We have built a series of tools --- map, trails, annotations and signposts --- based on a physical-world navigation metaphor. These tools have been in use for over a year. Our user study involved a controlled browse task and showed that users were able to get the same amount of work done with significantly less effort. Keywords information navigation, information foraging, interaction history, Web browsing INTRODUCTION Digital information has no history. It comes to us devoid of the patina that forms on physical objects as they are used. In the non-digital world we make extensive use of these traces to guide our actions, to make choices, and to find things of importance or interest. We call this area interaction history; that is, the records of the interactions of people and objects. Physical o...