@Techreport{ManevitzMalik1, Title = "Document Classification via Neural Networks Trained Exclusively with Positive Examples", author = "L. Manevitz and M. Yousef ", Institution = "Department of Computer Science, University of Haifa, Haifa", year = "2001", }, @TechReport{Scholkopf1, Author ="B. Sch{\"{o}}lkopf and J.C. Platt and J.Shawe-Taylor and A.J. Smola and R.C. Williamson", Title = "Estimating the Support of a High-Dimensional Distribution", Institution = "Microsoft Research,MSR-TR-99-87", Year ="1999"}, @ARTICLE{Burges, AUTHOR = "C.J.C Burges", TITLE = " A tutorial on support vector machines for pattern recognition", JOURNAL = "Data Mining and Knowledge Discovery ", YEAR = "1998", Volume = 2, } @INPROCEEDINGS{Dagan97, AUTHOR = "Ido Dagan and Yael Karov and Dan Roth", TITLE = "Mistake-driven learning in text categorization", BOOKTITLE = "Proceedings of EMNLP-97, 2nd Conference on Empirical Methods in Natural Language Processing", PUBLISHER = "Association for Computational Linguistics, Morristown, US", EDITOR = "Claire Cardie and Ralph Weischedel", YEAR = "1997", ADDRESS = "Providence, US", PAGES = "55--63", URL = "http://l2r.cs.uiuc.edu/~danr/Papers/categ.ps.gz", ABSTRACT = "Learning problems in the text processing domain often map the text to a space whose dimensions are the measured features of the text, e.g., its words. Three characteristic properties of this domain are (a) very high dimensionality, (b) both the learned concepts and the instances reside very sparsely in the feature space, and (c) a high variation in the number of active features in an instance. In this work we study three mistake-driven learning algorithms for a typical task of this nature - text categorization. We argue that these algorithms which categorize documents by learning a linear separator in the feature space have a few properties that make them ideal for this domain. We then show that a quantum leap in performance is achieved when we further modify the algorithms to better address some of the specific characteristics of the domain. In particular, we demonstrate (1) how variation in document length can be tolerated by either normalizing feature weights or by using negative weights, (2) the positive effect of applying a threshold range in training, (3) alternatives in considering feature frequency, and (4) the benefits of discarding features while training. Overall, we present an algorithm, a variation of Littlestone's Winnow, which performs significantly better than any other algorithm tested on this task using a similar feature set.", } @ARTICLE{Apte94, AUTHOR = "Apt\'{e}, Chidanand and Damerau, Fred J. and Weiss, Sholom M.", TITLE = "Automated learning of decision rules for text categorization", JOURNAL = "ACM Transactions on Information Systems", YEAR = "1994", NUMBER = "3", VOLUME = "12", PAGES = "233--251", URL = "http://www.research.ibm.com/dar/papers/pdf/tois94_with_cover.pdf", ABSTRACT = "We describe the results of extensive experiments using optimized rule-based induction methods on large document collections. The goal of these methods is to discover automatically classification patterns that can be used for general document categorization or personalized filtering of free text. Previous reports indicate that human-engineered rule-based systems, requiring many man-years of developmental efforts, have been successfully built to ``read'' documents and assign topics to them. We show that machine-generated decision rules appear comparable to human performance, while using the identical rule-based representation. In comparison with other machine-learning techniques, results on a key benchmark from the Reuters collection show a large gain in performance, from a previously reported 67% recall/precision breakeven point to 80.5%. In the context of a very high-dimensional feature space, several methodological alternatives are examined, including universal versus local dictionaries, and binary versus frequency related features.", } @INPROCEEDINGS{Yang99, AUTHOR = "Yiming Yang and Xin Liu", TITLE = "A re-examination of text categorization methods", BOOKTITLE = "Proceedings of SIGIR-99, 22nd ACM International Conference on Research and Development in Information Retrieval", EDITOR = "Marti A. Hearst and Fredric Gey and Richard Tong", PUBLISHER = "ACM Press, New York, US", ADDRESS = "Berkeley, US", YEAR = "1999", PAGES = "42--49", URL = "http://www.cs.cmu.edu/~yiming/papers.yy/sigir99.ps", ABSTRACT = "This paper reports a controlled study with statistical significance tests on five text categorization methods: the Support Vector Machines (SVM), a k-Nearest Neighbor (kNN) classifier, a neural network (NNet) approach, the Linear Least-squares Fit (LLSF) mapping and a Naive Bayes (NB) classifier. We focus on the robustness of these methods in dealing with a skewed category distribution, and their performance as function of the training-set category frequency. Our results show that SVM, kNN and LLSF significantly outperform NNet and NB when the number of positive training instances per category are small (less than ten), and that all the methods perform comparably when the categories are sufficiently common (over 300 instances).", } @INPROCEEDINGS{Schapire98, AUTHOR = "Schapire, Robert E. and Singer, Yoram and Singhal, Amit", TITLE = "Boosting and {R}occhio applied to text filtering", BOOKTITLE = "Proceedings of SIGIR-98, 21st ACM International Conference on Research and Development in Information Retrieval", EDITOR = "W. Bruce Croft and Alistair Moffat and Cornelis J. van Rijsbergen and Ross Wilkinson and Justin Zobel", PUBLISHER = "ACM Press, New York, US", YEAR = "1998", ADDRESS = "Melbourne, AU", PAGES = "215--223", URL = "http://www.research.att.com/~schapire/cgi-bin/uncompress-papers/SchapireSiSi98.ps", ABSTRACT = "We discuss two learning algorithms for text filtering: modified Rocchio and a boosting algorithm called AdaBoost. We show how both algorithms can be adapted to maximize any general utility matrix that associates cost (or gain) for each pair of machine prediction and correct label. We first show that AdaBoost significantly outperforms another highly effective text filtering algorithm. We then compare AdaBoost and Rocchio over three large text filtering tasks. Overall both algorithms are comparable and are quite effective. AdaBoost produces better classifiers than Rocchio when the training collection contains a very large number of relevant documents. However, on these tasks, Rocchio runs much faster than AdaBoost.", } @INPROCEEDINGS{Hull94, AUTHOR = "Hull, David A.", TITLE = "Improving text retrieval for the routing problem using latent semantic indexing", BOOKTITLE = "Proceedings of SIGIR-94, 17th ACM International Conference on Research and Development in Information Retrieval", EDITOR = "W. Bruce Croft and Cornelis J. van Rijsbergen", PUBLISHER = "Springer Verlag, Heidelberg, DE", YEAR = "1994", ADDRESS = "Dublin, IE", PAGES = "282--289", URL = "http://www.acm.org/pubs/articles/proceedings/ir/188490/p282-hull/p282-hull.pdf", ABSTRACT = "Latent Semantic Indexing (LSI) is a novel approach to information retrieval that attempts to model the underlying structure of term associations by transforming the traditional representation of documents as vectors of weighted term frequencies to a new coordinate space where both documents and terms are represented as linear combinations of underlying semantic factors. In previous research, LSI has produced a small improvement in retrieval performance. In this paper, we apply LSI to the routing task, which operates under the assumption that a sample of relevant and non-relevant documents is available to use in constructing the query. Once again, LSI slightly improves performance. However, when LSI is used is conduction with statistical classification, there is a dramatic improvement in performance.", } @ARTICLE{rocchio, AUTHOR = "J.J Jr Rocchio ", TITLE = "Relevance feedback in information retrieval", JOURNAL = "The SMART Retrieval System:Experiments in Automatic Document Processing", YEAR = "1971", NUMBER = "", VOLUME = "", PAGES = "313-323", URL = "", ABSTRACT = "", } @TechReport{JHF, Author ="J.H. Friedmann", Title = "Flexible metric nearest neigbor classification", Institution = "Nov,1994", } @INPROCEEDINGS{Lewis94, AUTHOR = "Lewis, David D. and Marc Ringuette", TITLE = "A comparison of two learning algorithms for text categorization", BOOKTITLE = "Proceedings of SDAIR-94, 3rd Annual Symposium on Document Analysis and Information Retrieval", PUBLISHER = "", EDITOR = "", YEAR = "1994", ADDRESS = "Las Vegas, US", PAGES = "81--93", URL = "http://www.research.att.com/\~{}lewis/papers/lewis94b.ps", ABSTRACT = "This paper examines the use of inductive learning to categorize natural language documents into predefined content categories. Categorization of text is of increasing importance in information retrieval and natural language processing systems. Previous research on automated text categorization has mixed machine learning and knowledge engineering methods, making it difficult to draw conclusions about the performance of particular methods. In this paper we present empirical results on the performance of a Bayesian classifier and a decision tree learning algorithm on two text categorization data sets. We find that both algorithms achieve reasonable performance and allow controlled tradeoffs between false positives and false negatives. The stepwise feature selection in the decision tree algorithm is particularly effective in dealing with the large feature sets common in text categorization. However, even this algorithm is aided by an initial prefiltering of features, confirming the results found by Almuallim and Dietterich on artificial data sets. We also demonstrate the impact of the time-varying nature of category definitions.", } @ARTICLE{chen-ng, AUTHOR = "H. Chen and T. Ng", TITLE = "An Algorithmic Approach to Concept Exploration in a Large Knowledge Network (Automatic Thesaurus Consultation): Symbolic Branch-and-Bound Search vs. Connectionist Hopfield Net Activation", JOURNAL = "Journal of the American Society for Information Science", YEAR = "1993", NUMBER = "5", VOLUME = "46", PAGES = "348-369", URL = "", ABSTRACT = "", } @ARTICLE{Macleod, AUTHOR = " K. MacLeod and W. Robertson", TITLE = "A Neural Algorithm for Document Clustering", JOURNAL = "Information Processing and Management", YEAR = "1991", NUMBER = "3", VOLUME = "27", PAGES = "337-346", URL = "", ABSTRACT = "", } @INPROCEEDINGS{lin-soergel-march, AUTHOR = "X. Lin, D. Soergel and G. Marchionini", TITLE = "A Self-Organizing Semantic Map for Information Retrieval", BOOKTITLE = "Proceedings of the 14th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval", YEAR = "1991", ADDRESS = "", PAGES = "262-269", PUBLISHER = "", URL = "", ABSTRACT = "", } @INPROCEEDINGS{wong-cai-yao, AUTHOR = "S.K.M. Wong,Y.J. Cai and Y.Y. Yao", TITLE = "Computation of Term Association by neural Network", BOOKTITLE = "SIGIR '93 Proceedings of the Sixteenth Annual International ACM SIGIR Conference on Research and Development in Information Retrieval", YEAR = " 1993", ADDRESS = "", PAGES = "", PUBLISHER = "", URL = "", ABSTRACT = "", } @ARTICLE{belew-rose, AUTHOR = "D. E. Rose and R. K. Belew", TITLE = "A connectionist and symbolic hybrid for improving legal reserach", JOURNAL = "Int. Journal of Man-Machine Studies", YEAR = " 1991", NUMBER = "1", VOLUME = "35", PAGES = "1-33", URL = "", ABSTRACT = "", } @INPROCEEDINGS{belew, AUTHOR = "R. K. Belew", TITLE = " Adaptive Information Retrieval", BOOKTITLE = " Proceedings of the Twelfth Annual International ACM/SIGIR Conference on Research and Development in Information Retrieval", EDITOR = " ", YEAR = " 1989", ADDRESS = "NY, NY", PAGES = "11-20", PUBLISHER = "", URL = "", ABSTRACT = "", } @INPROCEEDINGS{Mladenic99a, AUTHOR = "Dunja Mladeni{\'{c}} and Marko Grobelnik", TITLE = "Feature selection for unbalanced class distribution and Naive Bayes", BOOKTITLE = "Proceedings of ICML-99, 16th International Conference on Machine Learning", EDITOR = "Ivan Bratko and Saso Dzeroski", YEAR = "1999", ADDRESS = "Bled, SL", PAGES = "258--267", PUBLISHER = "Morgan Kaufmann Publishers, San Francisco, US", URL = "http://www-ai.ijs.si/DunjaMladenic/papers/PWW/pwwICML99Final.ps.gz", ABSTRACT = "This paper describes an approach to feature subset selection that takes into account problem specifics and learning algorithm characteristics. It is developed for the Naive Bayesian classifier applied on text data, since it combines well with the addressed learning problems. We focus on domains with many features that also have a highly unbalanced class distribution and asymmetric misclassification costs given only implicitly in the problem. By asymmetric misclassification costs we mean that one of the class values is the target class value for which we want to get predictions and we prefer false positive over false negative. Our example problem is automatic document categorization using machine learning, where we want to identify documents relevant for the selected category. Usually, only about 1%-10% of examples belong to the selected category. Our experimental comparison of eleven feature scoring measures show that considering domain and algorithm characteristics significantly improves the results of classification.", } @INPROCEEDINGS{Mladenic98b, AUTHOR = "Dunja Mladeni{\'{c}}", TITLE = "Feature subset selection in text learning", BOOKTITLE = "Proceedings of ECML-98, 10th European Conference on Machine Learning", SERIES = "Lecture Notes in Computer Science", NUMBER = "1398", PUBLISHER = "Springer Verlag, Heidelberg, DE", EDITOR = "Claire N{\'{e}}dellec and C{\'{e}}line Rouveirol", ADDRESS = "Chemnitz, DE", PAGES = "95--100", YEAR = "1998", URL = "http://www-ai.ijs.si/DunjaMladenic/papers/PWW/pwwECML98.ps.gz", ABSTRACT = "This paper describes several known and some new methods for feature subset selection on large text data. Experimental comparison given on real-world data collected from Web users shows that characteristics of the problem domain and machine learning algorithm should be considered when feature scoring measure is selected. Our problem domain consists of hyperlinks given in a form of small-documents represented with word vectors. In our learning experiments naive Bayesian classifier was used on text data. The best performance was achieved by the feature selection methods based on the feature scoring measure called Odds ratio that is known from information retrieval.", } @INPROCEEDINGS{Yang97, AUTHOR = "Yiming Yang and Jan O. Pedersen", TITLE = "A comparative study on feature selection in text categorization", BOOKTITLE = "Proceedings of ICML-97, 14th International Conference on Machine Learning", EDITOR = "Douglas H. Fisher", YEAR = "1997", ADDRESS = "Nashville, US", PAGES = "412--420", PUBLISHER = "Morgan Kaufmann Publishers, San Francisco, US", URL = "http://www.cs.cmu.edu/~yiming/papers.yy/ml97.ps", ABSTRACT = "This paper is a comparative study of feature selection methods in statistical learning of text categorization. The focus is on aggressive dimensionality reduction. Five methods were evaluated, including term selection based on document frequency (DF), information gain (IG), mutual information (MI), a 2 -test (CHI), and term strength (TS). We found IG and CHI most effective in our experiments. Using IG thresholding with a k-nearest neighbor classifier on the Reuters corpus, removal of up to 98% removal of unique terms actually yielded an improved classification accuracy (measured by average precision). DF thresholding performed similarly. Indeed we found strong correlations between the DF, IG and CHI values of a term. This suggests that DF thresholding, the simplest method with the lowest cost in computation, can be reliably used instead of IG or CHI when the computation of these measures are too expensive. TS compares favorably with the other methods with up to 50% vocabulary reduction but is not competitive at higher vocabulary reduction levels. In contrast, MI had relatively poor performance due to its bias towards favoring rare terms, and its sen sitivity to probability estimation errors.", } @MASTERSTHESIS{Tan00, AUTHOR = "Chade-Meng Tan", TITLE = "Finding and Using High Quality Word-Pairs for Enhanced Text Categorization", SCHOOL = "Department of Computer Science, University of California at Santa Barbara", ADDRESS = "Santa Barbara, US", YEAR = "2000", URL = "http://www.serve.com/cmtan/Meng/MS_Thesis/", ABSTRACT = "Text categorization is the task of automated assignment of natural language texts to predefined categories based on their content. Most text categorization systems use single words (unigrams) as features. A deceptively simple idea for enhancing text categorization is investigated here, an idea that has been previously shown not to work. It is to identify useful word pairs (bigrams) made up of adjacent unigrams. For example, to identify the bigram ``artificial+intelligence'' from the unigrams ``artificial'' and ``intelligence''. It is intuitively obvious that many bigrams describe concepts better than their component unigrams. We devised and refined an elegantly simple statistical algorithm that is able to identify high-quality bigrams, with only two passes of the training data, and using only simple parameters such as word occurrence counts and information gain. The bigrams it found, while very small in numbers, were able to substantially raise the quality of the feature set (45.8\% increase in average information gain for one dataset). The algorithm was run on 22 categories from 2 manually pre-classified datasets. One dataset is a collection of websites from the Science hierarchy on the {{\sc Yahoo!}}\ website (14,477 documents, 160,975 unique words), the other is the Reuters-21578 dataset of Reuters news articles (22,173 documents, 42,418 unique words). All categories we studied showed increases in precision/recall break-even (BE) point, with the highest increase measured at 27.6\%. The increase in mirco-averaged BE point for the Yahoo dataset was 6.8\% (11.9\% macro-averaged) and that for the Reuters dataset was 2.6\% (6.3\% macro-averaged). The Yahoo dataset also showed significant increase in the F1 measure (micro-average: 6.7\%, macro-average: 12.1\%). However, the Reuters dataset showed only insignificant increases in the same measure (micro-average: 0.7\%, macro-average: 3.0\%). Our study also revealed the main strength and weakness of our algorithm. It is most successful at increasing the number of correctly classified positive documents, but its main weakness is that it causes more negative documents to be classified incorrectly. This explains why it performed much better on the Yahoo-Science than the Reuters dataset, since the latter has a much higher recall and lower precision (ie, most of its positive documents are already correctly classified). Finally, we suggest a possible solution to this problem: adding a second stage to our current classifier that is specially trained to identify negative documents incorrectly classified by the first stage.", } @ARTICLE{Apte94, AUTHOR = "Apt\'{e}, Chidanand and Damerau, Fred J. and Weiss, Sholom M.", TITLE = "Automated learning of decision rules for text categorization", JOURNAL = "ACM Transactions on Information Systems", YEAR = "1994", NUMBER = "3", VOLUME = "12", PAGES = "233--251", URL = "http://www.research.ibm.com/dar/papers/pdf/tois94_with_cover.pdf", ABSTRACT = "We describe the results of extensive experiments using optimized rule-based induction methods on large document collections. The goal of these methods is to discover automatically classification patterns that can be used for general document categorization or personalized filtering of free text. Previous reports indicate that human-engineered rule-based systems, requiring many man-years of developmental efforts, have been successfully built to ``read'' documents and assign topics to them. We show that machine-generated decision rules appear comparable to human performance, while using the identical rule-based representation. In comparison with other machine-learning techniques, results on a key benchmark from the Reuters collection show a large gain in performance, from a previously reported 67% recall/precision breakeven point to 80.5%. In the context of a very high-dimensional feature space, several methodological alternatives are examined, including universal versus local dictionaries, and binary versus frequency related features.", } @INPROCEEDINGS{Joachims00, AUTHOR = "Thorsten Joachims", TITLE = "Estimating the Generalization Performance of a {SVM} Efficiently", BOOKTITLE = "Proceedings of ICML-00, 17th International Conference on Machine Learning", EDITOR = "", YEAR = "2000", ADDRESS = "Stanford, US", PAGES = "", PUBLISHER = "Morgan Kaufmann Publishers, San Francisco, US", URL = "http://www-ai.cs.uni-dortmund.de/DOKUMENTE/joachims_00a.pdf", ABSTRACT = "This paper proposes and analyzes an efficient and effective approach for estimating the generalization performance of a support vector machine (SVM) for text classification. Without any computation-intensive resampling, the new estimators are computationally much more efficient than cross-validation or bootstrapping. They can be computed at essentially no extra cost immediately after training a single SVM. Moreover, the estimators developed here address the special performance measures needed for evaluating text classifiers. They can be used not only to estimate the error rate, but also to estimate recall, precision, and F1. A theoretical analysis and experiments show that the new method can effectively estimate the performance of SVM text classifiers in an efficient way.", } @PHDTHESIS{DDL, AUTHOR = "D. D. Lewis", TITLE = "Representation and Learning in Information Retreival", SCHOOL = "Computer scince Dept. University of Massachussetts at Amherts ", YEAR = "Febrauary,1992"}, @InProceedings{LSCP, Author = "D.Lewis and R.E. Schapire and J.P. Callan and R.Papka", Title = " Training algorithms for Linear text Classifiers", booktitle = "Proceedings of the 19th Annual International ACM SIGIR Conference on research and development in information Retreival", year = "1996", pages = "298-306", publisher = "" } @ARTICLE{RHW, AUTHOR = "D. Rumelhard ,G. Hinton and R. Williams ", TITLE = " Learning internal representation by error propogation", JOURNAL = "Parallel distributed processing:Explorations in the Microstructure of Cognition ,MA:MIT Press,Cambridge" , YEAR = "1986", Volume ="" , pages = "" } @InProceedings{DPHS, Title = "Inductive learning algorithms and representation for text categorization", author = "S.T. Dumais and J. Platt and D. Heckerman and M. Sahami", booktitle = "Proceedings of the seventh International Conference on Information and Knowledge Management (CIKM'98)", year = "1998", pages = "148-155", publisher = "" } @InProceedings{TJ, Title = "Text categorization with support vector machines: Learning with many relevant features", author = "T. Joachims", booktitle = "Proceeding 10 European Conference on Machine Learning (ECML)", year = "1998", pages = "137-142", publisher = "Springer Verlag", URL="http://www-ai.cs.uni-dortmund.de/DOKIMENTE/Joachims\_97a.sp.gz", } , @BOOK{VNV, AUTHOR = "Vladimir N. Vapnik ", TITLE = "The nature of statistical learning theory", Publisher = "Springer", Address = "New York", Year = 1995}, @InProceedings{DP, Title = "Beyond independence: conditions for the optimality of the simple bayesian classifier", author = "P. Domingos and M. Pazzani", booktitle = "the Thirteenth international conference on machine learning ", year = "1996", publisher="Italy. Morgan Kaufmann", }, @ARTICLE{HM, AUTHOR = "R. Hummel and L. Manevitz", TITLE = " A Statistical Approach to the Representation of Uncertainty in Beliefs Using Spread of Opinions", JOURNAL = " IEEE Transactions on Systems, Man, and Cybernetics, Part A: Systems and Humans", YEAR = "1996", Volume = 3, pages = "378-384"} @InProceedings{KL, Title = "News{W}eeder:{L}earning to filter news", author = "K. Lang", booktitle = "Twelfth International Conference on Machine Learning", year = "1995", pages = "331-339", publisher="Lake Tahoe,CA", }, @PHDTHESIS{PDatta, AUTHOR = "P. Datta", TITLE = "Characteristic Concept Representations", SCHOOL = "University of California, Irvine", YEAR = "1997"}, @BOOK{Rij, AUTHOR = "C.J. van Rijsbergen", TITLE = "Information Retrieval", Publisher = "Butterworths", Address = "London, second edition", Year = 1979}, @InProceedings{Bos, Title = " Predicate Invention and Learning from Positive Examples Only", author = "H. Bostrom", booktitle = "Proc. of the Tenth European Conference on Machine Learning, Springer Verlag (1998)", year = "1998", pages="226-237", }, @InProceedings{DKR, Title = "Inductive learning algorithms and representations for text categorization", author = "S. Dumais, J. Platt and D. Heckerman", booktitle = "in Proceedings of ACM-CIKM98", year = "Nov. 1998", }, @Article{Mug, Author = "S. Muggleton", Title = "Learning from positive data", Journal = "Machine Learning", year =1999 }, @InProceedings{DKR, Title = "Mistake-Driven Learning in Text Categorization", author = "I. Dagan, Y. Karov and D. Roth", booktitle = "Advances in Classification Research vol. 8: Proceedings of the 8th ASIS SIG/CR Classification Research Workshop", year = "1998", pages="59-72", publisher = "Modford:New Jersey"}, @TechReport{YY, Author ="Y. Yang", Title = "An Evaluation of statistical approach to text categorization", Institution = "Carnegie Mellon University,CMU-CS-97-127", Year ="April 1997"}, @InProceedings{RS1, Title = "Combining Machine learning and heirarchical indexing structures for text categorization", author = " M.E. Ruiz and P. Srinivasan", booktitle = "Advances in Classification Research vol. 10: Proceedings of the 10th ASIS SIG/CR Classification Research Workshop", year = "1999", pages="", publisher = "Washington DC"}, @InProceedings{RS, Title = "Automatic Text Categorization Using Neural Networks", author = " M.E. Ruiz and P. Srinivasan", booktitle = "Advances in Classification Research vol. 8: Proceedings of the 8th ASIS SIG/CR Classification Research Workshop", year = "1998", pages="59-72", publisher = "Modford:New Jersey"}, @InProceedings{RDT, Title = " WebWatcher: A Learning Apprentice for the World Wide Web", author = "R. Armstrong, D. Freitag, T. Joachims and T.Mitchell", booktitle = "Working Notes of the AAAI Spring Symposium Series on Information Gathering from Distributed, Heterogeneous Environments", year = "1995", publisher = "CA:AAAI"}, @InProceedings{MY, Title = "Learning Information Retrieval Agents: Experiments with Automated Web Browsing", author = "M. Balabanovic and Y. Shoham", booktitle = "Working Notes of AAAI Spring Symposium Series on Information Gathering from Distributed, Heterogeneous Environments", year = "1995", publisher = "AAAI-Press"}, @InProceedings{MMB, Title = "Syskill \& {W}ebert:{I}dentifying Interesting Web Sites", author = "M. Pazzani and J. Muramatsu and D. Billsus", booktitle = "AAAI Conference 1996", pages = "54-61", year = "1996" }, @InProceedings{STKT, Title = "Creating and Order in Digital Libraries with Self-Organizing Maps", author = "S. Kaski,T. Honkela,K. Lagus and T. Kohonen", booktitle = "WCNN'96,World Congress on Neural Networks", year = "1996", pages = "814-817", publisher = "INNS Press"}, @TechReport{TSKT, Author ="T. Honkela,S. Kaski, K. Lagus and T. Kohonen", Title = " Newsgroup Exploration with WEBSOM Method and Browsing Interface", Institution = "Helsinki University of Technology, Laboratory of Computer and Information Science", Year ="1996", }, @BOOK{TK, AUTHOR = "T. Kohonen", TITLE = "Self-Organization Maps", Publisher = "Springer-Verlag", Address = "Berlin", Year = 1995}, @BOOK{Sw, AUTHOR = " K. Swingler", TITLE = "Applying Neural Networks: a Practical Guide", Publisher = " Academic Press", Address = " New York", Year = 1996}, @BOOK{Ko, AUTHOR = "T. Kohonen", TITLE = "Self-Organization and Associative Memory, Second Edition", Publisher = "Springer-Verlag", Address = "Berlin", Year = 1988} @mastersthesis{CQ, AUTHOR = "C. Quek", TITLE = "Classification of World Wide Web Documents", SCHOOL = " School of Computer Science Camegie Mellon University", YEAR=1997 }, @TechReport{EIT, Author ="E. Bloedron, I. Mani and T. MacMillan", Title = "Representational Issues in Machine Learning of User Profiles", Institution = "Artificial Intelligence Technical Center, The MITRE Corporation, Z401" }, @mastersthesis{BS, AUTHOR = "B.D. Sheth", TITLE = "A Learning Approach to Personalized Information Filtering", SCHOOL = "Massachusetts Institute of Technology", YEAR = 1994}, @BOOK{GM, AUTHOR = "G. Salton and M.J. McGill", TITLE = "Introduction to Modern Information Retrieval", Publisher = "McGraw-Hill", Address = "Berlin", Year = 1983}, @BOOK{TM, AUTHOR = "T. Mitchell", TITLE = "Machine Learning", Publisher = "McGraw-Hill", Year =1996 }, @InProceedings{OMP, Title = "Adaptive Web Sites:an AI Challenge", author = "M. Perkowitz and O. Etzioni", booktitle = "15th Int. Conf. Autonomous Agents", year = "1997" }, @InProceedings{MOE, Title = "Adaptive Web Sites:Automatically Learning from User Access Patterns", author = "M. Perkowitz and O. Etzioni", booktitle = "Proceedings of the Sixth Int. WWW Conference", year = "1997" }, @PHDTHESIS{SK, AUTHOR = "S. Kaski", TITLE = "Data Exploration Using Self-Organizing Maps", SCHOOL = "Helsinki University of Technology, Neural Networks Research Center, Finland ", YEAR = "1997"}, @Article{CAJ, Author = "C.V. Goldman, A. Langer and J.S. Rosenschein", Title = "MUSAG: An Agent that Learns What you Mean", Journal = "Applied Artificial Intelligence", year =1997 }, @Article{JASP, Author = "J.F. Arcand and S. Pelletier", Title = "Agents: From Theoretical Foundations to Practical Applications", Journal = "AI, Canadian Artificial Intelligence", year =1996 }, @Article{TSE, Author = "T. Selker", Title = "COACH: A Teaching Agent That Learns", Journal = "Communications of the ACM", month="july", year = 1994}, @InProceedings{WPW, Author =" E. Weiner and J.O. Pedersen and A.S. Weigned", Title = " A neural network approach to topic spotting", booktitle="4th Symposium on document analysiss and information retreival, las vegas", pages="317-332", year=1995 }, @InProceedings{PPR, Title = "How to Build Modeling Agents to Support Web Searchers", author = "P.P. Maglio and R. Barrett", booktitle = "Proceeding of the Sixth International Conference, UM97. Vienna", year = "1997", publisher = "New York:Springer Wien New York"}, @InProceedings{RPD, Title = "How to Personalize the Web", author = "R. Barret, P.P. Maglio and D.C. Kellem", booktitle = "Proceeding of the Conference on Human Factors in Computer System (CHI 97')", year = "1997", publisher = "NY:ACM Press"}, @Article{HL, Author = "H. Lieberman", Title = "Letizia: An Agent that Assists Web Browsing", Journal = "International Joint Conference on Artificial Intelligence", year = 1995}, @InProceedings{HLI, Title = "Autonomous Interface Agents", author = "H. Lieberman", booktitle = "Proceeding of the ACM Conference on Computer and Human Interface, CHI97, Atlanat, Georgia", year = "1997" }, @Article{MD, Author = "M. Pazzani and D. Billsus", Title = "Learning and Revising User Profiles: The Identification of Interesting Web Sites", Journal = "Machine Learning ", Volume = 27, pages = "313-331", year = 1997}, @TechReport{AJH, Author ="A. Jennings and H. Higuchi", Title = "A Personal News Service based on a User Model Neural Network", Institution = "Kansai Advanced Research Center, Communications Research Laboratory", Address = "Iwaoka, Kobe, Japan" }, @Article{QC, Author = "Q. Chen", Title = "Modeling A User's Domain Knowledge with Neural Networks", Journal = "Journal of Human-Computer Interaction" }, @InProceedings{NCM, Title = "A Novelty Detection Approach to Classification", author = "N. Japkowicz and C. Myers and M. Gluck", booktitle = "Proceeding of the Fourteenth International Conference On Artificial Intelligence", year = "1995", pages = "518-523", publisher = "Montreal, Canada"}, @TechReport{TJO, Author ="T. Joachims", Title = "A Probabilistic Analysis of the {R}occhio Algorithm with {TF-IDF} for Text Categorization", Number = "CMU-CS-96-118", Institution = "School of Computer Science, Carnegie Mellon University, Pittsburgh", Year ="1996"}, @Article{RDL, Author = "D. Lewis", Title = "Reuters-21578 Text Categorization Test Collection", Journal = "http://www.research.att.com/\~{}lewis", year ="1997" }, @BOOK{RZ, Author = "J.S. Rosenschein and G. Zlotkin", Title = "Rules of Encounter", Publisher = "MIT Press ", Address = "London, England", Year = 1994}, @BOOK{RMS, Author = "H. Ritter, T. Martinetz and K. Schulten", Title = "Neural Computation and Self-Organizing Maps", Publisher = "ADDISON-WESLEY ", Year = 1992}, @INCOLLECTION{CMZ, AUTHOR = "G.W. Cottrell, P. Munro and D. Zipser", Title = "Image Compression by Back Propagation: an Example of Extensional Programming", booktitle = "Advances in Cognitive Science", volume= 3, editor = "N.E. Sharkey", publisher = "Ablex", year = "1988"}, @Article{KOTYPEWRITER, Author = "T. Kohonen ", Title = "The 'neural' Phonetic Typewriter", Journal = "Computer", Number = 21, pages = "11-22", year = 1988}, @Article{MP, Author = "M. Porter", Title = "An Algorithm for Suffix Stripping", Journal = "program", Number = 14, pages = "130-137", year = 1980}, @INCOLLECTION{MG, AUTHOR = "L. Manevitz and D. Givoli", Title = "The Finite Element Method and Soft Computing", booktitle = "3rd On-line World Conf. on Soft Computing in Engineering Design and Manufacturing(WSC3)", editor = "R. Roy, 1999", publisher = "Springer-Verlag", year = "June 1998"}, @BOOK{Rich, Author = "Elaine Rich and Kevin Knight", Title = "Artificial Intelligence, 2nd edition", Publisher = " McGraw Hill", Address = "New York", Year = 1991}, @Book{Winston, Author = " P. Winston", Title = "Artificial Intelligence, 3rd edition", Publisher = "Addison-Wesley", Address = "New York", Year = 1992}, @Article{MSM, Author = " M. Meltser, M. Shoham and L. Manevitz", Title = " Approximating Functions by Neural Networks: A Constructive Solution in the Uniform Norm", Journal = "Neural Networks", Volume = 9, Number = 6, pages = "965-978", year = 1996}, @BOOK{PDP, Author = " D.E. Rumelhart and J. L. McClelland", Title = " Parallel Distributed Processing", Publisher = " MIT Press", Address = "Cambridge, MA", Year = 1986}, @BOOK{SDM, Author = "Pentti Kanerva", Title = "Sparse Distributed Memory", Publisher = "MIT Press", Address = "cambridge, MA", Year = 1988}, @TechReport{TSDM, Author =" L. Manevitz", Title = " The 'Sense of Time' in Associative Memories", Institution = "The University of Haifa", Year ="1996"}, @ARTICLE{Cy, Author = " G. Cybenko", Title = " Approximation by superpositions of a sigmoidal function", Journal = "Mathematics of Control Signals and Systems", pages = "303 - 314", Volume = 2, Year = 1989}, @BOOK{GJ, AUTHOR = "M.R. Garey and D. S. Johnson", Title = "Computers and Intractability: a Guide to the Theory of NP-Completeness", Publisher = " Freeman", Year = 1979}, @ARTICLE{ManLetters, AUTHOR = "L. M. Manevitz", Title = "Interweaving Kohonen Maps of Different Dimensions to Handle Measure Zero Constraints on Topological Mappings", Journal = "Neuroprocessing Letters", pages = "to appear", year = "1997"}, @INCOLLECTION{SCE, AUTHOR = "J.D. Schaffer, R. A. Caruna and L. J. Eshelman", Title = "Using genetic search to exploit the emergent behavior of neural networks", booktitle = "Emergent Computation", pages = "244-248", editor = "S. Forrest", publisher = "Morgan Kaufmann", year = "1990"}, @InProceedings{Glovetalk, author = "Sidney Fels and Geoffrey Hinton", title = "GloveTalk{II}: An Adaptive Gesture-to-Formant Interface", booktitle = "Proceedings of ACM CHI'95 Conference on Human Factors in Computing Systems", pages = "456--463", year = "1995"}, @ARTICLE{Wid, Author = "B. Widrow and M. Lehr", Title = "30 years of adaptive neural networks: perceptron, madaline, and backpropagation", journal = "Proceedings of the IEEE", volume = 78, number = 9, pages = "1415-1442", year = 1990}, @InProceedings{HallR, Author = "Lawrence O. Hall and Steve G. Romaniuk", Title = "A Hybrid Connectionist, Symbolic Learning System", pages = "783--788", editor = "William Dietterich, Tom; Swartout", booktitle = "Proceedings of the 8th National Conference on Artificial Intelligence", year = "1990", publisher = "MIT Press" }, @InProceedings{LIS, title = "Neural network-assisted Japanese-English machine translation system", author = "T. Law and H. Itoh and H. Seki", booktitle = "Proceedings of 1993 International Joint Conference on Neural Networks. Part 3 (of 3) Nagoya, Jpn Oct 25-25", year = "1993", volume = "3", pages = "2905--2908", publisher = "IEEE"}, @ARTICLE{SST, Author = " A. Schaerf, Y. Shoham and M. Tennenholtz", Title = "Adaptive Load Balancing: A Study in Multi-Agent Learning", Journal = " Journal of Artificial Intelligence Research", Volume = 2, pages = "475-500", year = "1995"}, @BOOK{Yafuzzy, AUTHOR = "R. R. Yager", Title = "Essentials of Fuzzy Modeling and Control", publisher = "J. Wiley", year = 1994} @BOOK{Yasoft1, AUTHOR = "R. R. Yager and B. Bouchon-Meunier", Title = "Fuzzy Logic and Soft Computing", Publisher = "World Scientific", year = 1995}, @BOOK{Yasoft2, Editor = "R. R. Yager and Lofti A. Zadeh", Title = "Fuzzy Sets, Neural Networks, and Soft Computing", Publisher = "World Scientific", year = 1995}, @BOOK{HrW, Editor = "F. Hayes-Roth, D.A. Waterman and D. Lenat", Title = "Building Expert Systems", Publisher = "Addison-Wesley", year = 1983}, @BOOK{WS, Author = "B. Widrow and S. Stearns", Title = "Adaptive signal Processing", Publisher = "Prentice Hall", year = 1985}, @BOOK{Ros, Author = "J. Rosenschein", Title = "Rules of Encounter", Publisher = "MIT Press", year = 1994}, @BOOK{Go, Author = "D. Goldberg", Title = "Genetic Algorithms", Publisher = "Addison-Wesley", year = 1989}, @BOOK{CO, AUTHOR = "G.F. Carey and J.T. Oden", TITLE = "Finite Elements, Vol. III: Computational Aspects", Publisher = "Prentice-Hall", Address = "Englewood Cliffs, NJ", Year = 1984} @BOOK{Ko, AUTHOR = "T. Kohonen", TITLE = "Self-Organization and Associative Memory, Second Edition", Publisher = "Springer-Verlag", Address = "Berlin", Year = 1988} @TECHREPORT{TI, AUTHOR = "T. Tzvi and E. Iaakov", TITLE = "Towards Real-Time Self-Organizing Network with Parallel and Noisy Inputs", Institution = "Hebrew University", Type = "Preprint", Year = 1994} @ARTICLE{Fr, AUTHOR = "B. Fritzke", Title = "Growing Cell Structures - A self-organizing network for unsupervised and supervised learning", JOURNAL = "Neural Neworks", Volume = 7, Pages = "1441-1460", YEAR = 1994} @BOOK{Jo, AUTHOR = " R.E. Jones", TITLE = "A Self-Organizing Mesh Generation Program", Publisher = "ASME Publication 74-PVP-13", Year = 1974} @ARTICLE{TTMW, AUTHOR = "F.C. Thames J.F. Thompson C.W. Mastin and R.L. Walker", TITLE = "Numerical Solutions for Viscous and Potential Flow about Arbitrary Two-Dimensional Bodies Using Body-Fitted Coordinate Systems", JOURNAL = "J. Comput. Phys. ", YEAR = "1977 ", Volume = 24 , pages = "245-273"} @ARTICLE{MMG, AUTHOR = "L. Manevitz D. Givoli and M. Margi", TITLE = "Heuristic finite element node numbering", JOURNAL = "Computing Systems in Engineering", YEAR = "1993", Volume = 4, Pages = "159-168"} @TECHREPORT{MYGtech, AUTHOR = "L. Manevitz Malik Yousef and D. Givoli", TITLE ="Finite Element Mesh Generation Using Self-Organizing Neural Networks", INSTITUTION = "University of Haifa", YEAR = "1995"} @ARTICLE{MYG, AUTHOR = "L. Manevitz Malik Yousef and D. Givoli", TITLE ="Finite Element Mesh Generation Using Self-Organizing Neural Networks", JOURNAL = "Microcomputers in Civil Engineering", Volume = 12, Number = 4, YEAR = "1997"}, @ARTICLE{Rh, AUTHOR = "D. Rhynsburger", TITLE = "Analytic Delineation of Thiessen Polygons", JOURNAL = "Geogr. Anal. ", YEAR = "1973", Volume = 5, pages = "133-144"} @INCOLLECTION{SF, AUTHOR = "M.S Shephard and P.M. Finnigan", TITLE = "Towards Automatic Model Generation", BOOKTITLE = "State of the Art Surveys on Computational Mechanics", EDITOR = "A.K. Noor and T.J. Oden", PUBLISHER = " ASME", YEAR = "1989", pages = "335-366"} @INCOLLECTION{Th, AUTHOR = "F. Thomasset", TITLE ="Appendix to Navier-Stokes Problems", BOOKTITLE="Navier-Stokes Problems", EDITOR = "R. Temam", PUBLISHER =" North Holland", YEAR = "1977"} @PHDTHESIS{Re, AUTHOR = " R. Renka", TITLE = "Triangulation and Bivariate Interpolation for Irregularly Distributed Data Points", SCHOOL = " University of Texas at Austin", YEAR = "1981"} @BOOK{Ba, AUTHOR = "R.E. Bank", TITLE = "PLTMG: A Software Package for Solving Elliptic Partial Differential Equations", Publisher = "SIAM publications", Address = "Philadelphia", YEAR = " 1994"} @BOOK{PLTMG, AUTHOR = "R.E. Bank", TITLE = "PLTMG: A Software Package for Solving Elliptic Partial Differential Equations", Publisher = "SIAM publications", Address = "Philadelphia", YEAR = " 1994"} @BOOK{Hu, AUTHOR = "T.J.R. Hughes", TITLE = "The Finite Element Method", Publisher = "Prentice-Hall", Address = "New York", Year = "1987" } @BOOK{Ge, AUTHOR = " P.L. George", TITLE = "Automatic Mesh Generation", Publisher = " Wiley", Address = "Chichester, UK", YEAR = " 1991"} @BOOK{KS, AUTHOR = " P. Knupp and S. Steinberg", TITLE = "Fundamentals of Grid Generation", Publisher = "CRC Press", Address = "Boca Raton, FL", YEAR = "1993"} @mastersthesis{Ma, AUTHOR = "Malik Yousef", Title = " Automatic Mesh Generation Using Self-Organizing Neural Networks", School = "University of Haifa", Year = " 1996", Note = "in Hebrew"} @BOOK{CO, AUTHOR = "G.F. Carey and J.T. Oden", TITLE = "Finite Elements, Vol. III: Computational Aspects", Publisher = "Prentice-Hall", Address = "Englewood Cliffs, NJ", Year = 1984} @BOOK{Ko, AUTHOR = "T. Kohonen", TITLE = "Self-Organization and Associative Memory, Second Edition", Publisher = "Springer-Verlag", Address = "Berlin", Year = 1988} @TECHREPORT{TI, AUTHOR = "T. Tzvi and E. Iaakov", TITLE = "Towards Real-Time Self-Organizing Network with Parallel and Noisy Inputs", Institution = "Hebrew University", Type = "Preprint", Year = 1994} @ARTICLE{Fr, AUTHOR = "B. Fritzke", Title = "Growing Cell Structures - A self-organizing network for unsupervised and supervised learning", JOURNAL = "Neural Neworks", Volume = 7, Pages = "1441-1460", YEAR = 1994} @BOOK{Jo, AUTHOR = " R.E. Jones", TITLE = "A Self-Organizing Mesh Generation Program", Publisher = "ASME Publication 74-PVP-13", Year = 1974} @ARTICLE{TTMW, AUTHOR = "F.C. Thames J.F. Thompson C.W. Mastin and R.L. Walker", TITLE = "Numerical Solutions for Viscous and Potential Flow about Arbitrary Two-Dimensional Bodies Using Body-Fitted Coordinate Systems", JOURNAL = "J. Comput. Phys. ", YEAR = "1977 ", Volume = 24 , pages = "245-273"} @ARTICLE{MMG, AUTHOR = "L. Manevitz D. Givoli and M. Margi", TITLE = "Heuristic finite element node numbering", JOURNAL = "Computing Systems in Engineering", YEAR = "1993", Volume = 4, Pages = "159-168"} @ARTICLE{Rh, AUTHOR = "D. Rhynsburger", TITLE = "Analytic Delineation of Thiessen Polygons", JOURNAL = "Geogr. Anal. ", YEAR = "1973", Volume = 5, pages = "133-144"} @INCOLLECTION{SF, AUTHOR = "M.S Shephard and P.M. Finnigan", TITLE = "Towards Automatic Model Generation", BOOKTITLE = "State of the Art Surveys on Computational Mechanics", EDITOR = "A.K. Noor and T.J. Oden", PUBLISHER = " ASME", YEAR = "1989", pages = "335-366"} @INCOLLECTION{Th, AUTHOR = "F. Thomasset", TITLE ="Appendix to Navier-Stokes Problems", BOOKTITLE="Navier-Stokes Problems", EDITOR = "R. Temam", PUBLISHER =" North Holland", YEAR = "1977"} @PHDTHESIS{Re, AUTHOR = " R. Renka", TITLE = "Triangulation and Bivariate Interpolation for Irregularly Distributed Data Points", SCHOOL = " University of Texas at Austin", YEAR = "1981"} @BOOK{Ba, AUTHOR = "R.E. Bank", TITLE = "PLTMG: A Software Package for Solving Elliptic Partial Differential Equations", Publisher = "SIAM publications", Address = "Philadelphia", YEAR = " 1994"} @BOOK{Hu, AUTHOR = "T.J.R. Hughes", TITLE = "The Finite Element Method", Publisher = "Prentice-Hall", Address = "New York", Year = "1987" } @BOOK{Ge, AUTHOR = " P.L. George", TITLE = "Automatic Mesh Generation", Publisher = " Wiley", Address = "Chichester, UK", YEAR = " 1991"} @BOOK{KS, AUTHOR = " P. Knupp and S. Steinberg", TITLE = "Fundamentals of Grid Generation", Publisher = "CRC Press", Address = "Boca Raton, FL", YEAR = "1993"} @mastersthesis{Ma, AUTHOR = "Malik Yousef", Title = " Automatic Mesh Generation Using Self-Organizing Neural Networks", School = "University of Haifa", Year = " 1996", Note = "in Hebrew"}