The automated categorization (or classification) of texts into predefined categories has witnessed a booming interest in the last ten years, due to the increased availability of documents in digital form and the ensuing need to organize them. In the research community the dominant approach to this problem is based on machine learning techniques: a general inductive process automatically builds a classifier by learning, from a set of preclassified documents, the characteristics of the categories. The advantages of this approach over the knowledge engineering approach (consisting in the manual definition of a classifier by domain experts) are a very good effectiveness, considerable savings in terms of expert labor power, and straightforward portability to different domains. This survey discusses the main approaches to text categorization that fall within the machine learning paradigm. We will discuss in detail issues pertaining to three different problems, namely document representation, classifier construction, and classifier evaluation.

Abstract not found

We propose a novel approach for categorizing text documents based on the use of a special kernel. The kernel is an inner product in the feature space generated by all subsequences of length k. A subsequence is any ordered sequence of k characters occurring in the text though not necessarily contiguously. The subsequences are weighted by anexponentially decaying factor of their full length in the text, hence emphasising those occurrences that are close to contiguous. A direct computation of this feature vector would involve a prohibitive amount of computation even for modest values of k, since the dimension of the feature space grows exponentially with k. The paper describes how despite this fact the inner product can be e ciently evaluated by a dynamic programming technique. Experimental comparisons of the performance of the kernel compared with a standard word feature space kernel Joachims (1998) show positive results on modestly sized datasets. The case of contiguous subsequences is also considered for comparison with the subsequences kernel with di erent decay factors. For larger documents and datasets the paper introduces an approximation technique that is shown to deliver good approximations e ciently for large datasets.

The optimal settings of retrieval parameters often depend on both the document collection and the query, and are usually found through empirical tuning. In this paper, we propose a family of two-stage language models for information retrieval that explicitly captures the different influences of the query and document collection on the optimal settings of retrieval parameters. As a special case, we present a two-stage smoothing method that allows us to estimate the smoothing parameters completely automatically. In the first stage, the document language model is smoothed using a Dirichlet prior with the collection language model as the reference model. In the second stage, the smoothed document language model is further interpolated with a query background language model. We propose a leave-one-out method for estimating the Dirichlet parameter of the first stage, and the use of document mixture models for estimating the interpolation parameter of the second stage. Evaluation on five different databases and four types of queries indicates that the twostage smoothing method with the proposed parameter estimation methods consistently gives retrieval performance that is close to— or better than—the best results achieved using a single smoothing method and exhaustive parameter search on the test data.

The amount of services and deployed software agents in the most famous offspring of the Internet, the World Wide Web, is exponentially increasing. In addition, the Internet is an open environment, where information sources, communication links and agents themselves may appear and disappear unpredictably.Thus, an effective, automated search and selection of relevant services or agents is essential for human users and agents as well. We distinguish three general agent categories in the Cyberspace, serviceproviders, servicerequester, and middle agents. Service providers provide some type of service, such as finding information, or performing some particular domain specific problem solving. Requester agents need provider agents to perform some service for them. Agents that help locate others are called middle agents [2]. Matchmaking is the process of finding an appropriate provider for a requester through a middl...

Abstract not found

This paper describes an experimental comparison of seven different learning algorithms on the problem of learning to disambiguate the meaning of a word from context. The algorithms tested include statistical, neural-network, decision-tree, rule-based, and case-based classification techniques. The specific problem tested involves disambiguating six senses of the word "line" using the words in the current and proceeding sentence as context. The statistical and neural-network methods perform the best on this particular problem and we discuss a potential reason for this ob- served difference. We also discuss the role of bias in machine ]earning and its importance in explaining performance differences observed on specific problems.

Our central claim is that user interactions with everyday productivity applications (e.g., word processors, Web browsers, etc.) provide rich contextual information that can be leveraged to support just-in-time access to task-relevant information. We discuss the requirements for such systems, and develop a general architecture for systems of this type. As evidence for our claim, we present Watson, a system which gathers contextual information in the form of the text of the document the user is manipulating in order to proactively retrieve documents from distributed information repositories. We close by describing the results of several experiments with Watson, which show it consistently provides useful information to its users.

The combination of different text representations and search strategies has become a standard technique for improving the effectiveness of information retrieval. Combination, for example, has been studied extensively in the TREC evaluations and is the basis of the “meta-search” engines used on the Web. This paper examines the development of this technique, including both experimental results and the retrieval models that have been proposed as formal frameworks for combination. We show that combining approaches for information retrieval can be modeled as combining the outputs of multiple classifiers based on one or more representations, and that this simple model can provide explanations for many of the experimental results. We also show that this view of combination is very similar to the inference net model, and that a new approach to retrieval based on language models supports combination and can be integrated with the inference net model.

In this paper, we introduce a class of systems called Information Management Assistants (IMAs). IMAs automatically discover related material on behalf of the user by serving as an intermediary between the user and information retrieval systems. IMAs observe users interact with everyday applications and then anticipate their information needs using a model of the task at hand. IMAs then automatically fulfill these needs using the text of the document the user is manipulating and a knowledge of how to form queries to traditional information retrieval systems (e.g., Internet search engines, abstract databases, etc.). IMAs automatically query information systems on behalf of users as well as provide an interface by which the user can pose queries explicitly. Because IMAs are aware of the user's task, they can augment their explicit query with terms representative of the context of this task. In this way, IMAs provide a framework for bringing implicit task context to bear on servicing expli...