There are many applications in which it is desirable to order rather than classify instances. Here we consider the problem of learning how to order, given feedback in the form of preference judgments, i.e., statements to the effect that one instance should be ranked ahead of another. We outline a two-stage approach in which one first learns by conventional means a preference function, of the form PREF(u; v), which indicates whether it is advisable to rank u before v. New instances are then ordered so as to maximize agreements with the learned preference function. We show that the problem of finding the ordering that agrees best with a preference function is NP-complete, even under very restrictive assumptions. Nevertheless, we describe a simple greedy algorithm that is guaranteed to find a good approximation. We then discuss an on-line learning algorithm, based on the "Hedge" algorithm, for finding a good linear combination of ranking "experts." We use the ordering algorith...

Recommender systems improve access to relevant products and information by making personalized suggestions based on previous examples of a user's likes and dislikes. Most existing recommender systems use collaborative filtering methods that base recommendations on other users' preferences. By contrast, content-based methods use information about an item itself to make suggestions. This approach has the advantage of being able to recommend previously unrated items to users with unique interests and to provide explanations for its recommendations. We describe a content-based book recommending system that utilizes information extraction and a machine-learning algorithm for text categorization. Initial experimental results demonstrate that this approach can produce accurate recommendations.

We describe the use of a recommender system to enable continuous knowledge acquisition and individualized tutoring of application software across an organization. Installing such systems will result in the capture of evolving expertise and in organization-wide learning (OWL). We present the results of a year-long naturalistic inquiry into an application's usage patterns, based on logging users' actions. We analyze the data to develop user models, individualized expert models, and instructional indicators. We show how this information is used to recommend learning tips to users. Keywords Recommender system, organization-wide learning, OWL, individualized instruction, agent, instrumentation, logging. Introduction New Workplace Technologies Enable New Learning Methods In the last decade, an enormous change has taken place in the workplace: there is a PC on every desk, and much office work is performed in the medium of software. Mastering one's software - at least the portion of...

Information technology has recently become the medium in which much professional ofce work is performed.This change offers an unprecedented opportunity to observe and record exactly how that work is performed.We describe our observation and logging processes and present an overview of the results of our long-term observations of a number of users of one desktop application. We then present our method of providing individualized instruction to each user by employing a new kind of user model and a new kind of expert model.The user model is based on observing the individual's behavior in a natural environment, while the expert model is based on pooling the knowledge of numerous individuals. Individualized instructional topics are selected by comparing an individual's knowledge to the pooled knowledge of her peers.

A contemporary approach to describing and theorizing about joint human endeavor is to posit “knowledge in common ” as a basis for awareness and coordination. Recent analysis has identified weaknesses in this approach even as it is typically employed in relatively simple task contexts. We suggest that in realistically complex circumstances, people share activities and not merely concepts. We describe a framework for understanding joint endeavor in terms of four facets of activity awareness: common ground, communities of practice, social capital, and human development. We illustrate the sort of analysis we favor with a scenario from emergency management, and consider implications and future directions for system design and empirical methods. 1.

Content-based recommender systems suggest documents, items, and services to users based on learning a profile of the user from rated examples containing information about the given items. Text categorization methods are very useful for this task but generally rely on unstructured text. We have developed a bookrecommending system that utilizes semi-structured information about items gathered from the web using simple information extraction techniques. Initial experimental results demonstrate that this approach can produce fairly accurate recommendations.

collaborates with a workgroup to increase information nugget sharing among the group. Information nuggets are small amounts of self-contained information, such as the URL of an interesting news article, a book title, or the time and location of a local art event. NuggetMine and the workgroup work together to build, maintain, and utilize a repository---or "mine"---of information nuggets. Group members submit nuggets to NuggetMine, which organizes and augments the submitted nuggets and provides a desktop interface to each group member. This interface makes it easy for group members to submit nuggets, view nuggets, and explore the mine. NuggetMine distributes the tasks necessary to share nuggets between it and the workgroup so as to best utilize the skills of each collaborator. In this paper, we describe the NuggetMine application and interface and present a pilot study of the application.

Author: Paul N. Bennett Title: Text Categorization Through Probabilistic Learning: Applications to Recommender Systems Supervising Professor: Raymond J. Mooney, Ph.D. With the growth of the World Wide Web, recommender systems have received an increasing amount of attention. Many recommender systems in use today are based on collaborative filtering. This project has focused on LIBRA, a content-based book recommending system. By utilizing text categorization methods and the information available for each book, the system determines a user profile which is used as the basis of recommendations made to the user. Instead of the bagof -words approach used in many other statistical text categorization approaches, LIBRA parses each text sample into a semi-structured representation. We have used standard Machine Learning techniques to analyze the performance of several algorithms on this learning task. In addition, we analyze the utility of several methods of feature construction and selection (...

“Free speech ” can mean not only face-to-face communication, but also expression embodied in the media of newspapers, books, television, film and so forth. Many of these media constitute public “spaces. ” With the introduction of each new public space, the theories and practices of “speech ” and “conversation ” are affected and extended. This article concerns a philosophical study of and artistic-design approach to some of the new, electronic, public spaces of the Internet and the forms of “speech, ” “conversation ” and dialectics practiced in these new spaces. The new electronic spaces that I am interested in have the following characteristics in common: • They are large. Many server sites now support interchanges between hundreds and thousands of people. Usenet newsgroups and large listservs are the most common of such

This research addresses the problem of keeping the skills of information technology (IT) users up to date when both information technology and job tasks are evolving rapidly. The approach we have taken is to use recommender system technology to pass experience from an immediate community of information technology users to the individual who is trying to decide which IT functionality to learn next. User profiles are constructed automatically from implicit input. We compare each user's profile to the pooled knowledge of their peers to identify gaps in their knowledge and to determine the most useful commands at the boundaries of their knowledge.