H. Ng, W. Goh, and K. Low. "Feature selection, perceptron learning, and a usability case study for text categorization". In Proc. 20th Int. ACM SIGIR Conf. on Research and Development in Information Retrieval (SIGIR97), pages 67--73, 1997.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....learn rules have been proposed for text categorisation. Most of these techniques employ the bag of words or word vector space feature representation [30] where each word in the text document corresponds to a single feature. A learning algorithm such as decision trees [3] neural networks [25], Bayesian probabilistic approaches [23] 42] or support vector machines [18] is then used to classify the text document. de Vel [9] studied the comparative performance of text document categorisation algorithms using the Naive Bayes, Support Vector Machines, multi layer Perceptron and k NN ....

H. Ng, W. Goh, and K. Low. "Feature selection, perceptron learning, and a usability case study for text categorization". In Proc. 20th Int. ACM SIGIR Conf. on Research and Development in Information Retrieval (SIGIR97), pages 67--73, 1997.

....learn rules have been proposed for text categorisation. Most of these techniques employ the bag of words or word vector space feature representation [29] where each word in the text document corresponds to a single feature. A learning algorithm such as decision trees [2] neural networks [24], Bayesian probabilistic approaches [22] 40] or support vector machines [17] is then used to classify the text document. de Vel [8] studied the comparative performance of text document categorisation algorithms using the Naive Bayes, Support Vector Machines, multi layer Perceptron and k NN ....

H. Ng, W. Goh, and K. Low. "Feature selection, perceptron learning, and a usability case study for text categorization". In Proc. 20th Int. ACM SIGIR Conf. on Research and Development in Information Retrieval (SIGIR97), pages 67--73, 1997.

....is a combinatorial feature selection algorithm. When possible, we use the Sequential Floating Backward Selection (SFBS) algorithm [20] Unfortu2 nately, we find that SFBS is not feasible for feature sets of more than about 110 features. Ng, et al., have previously reported a limit of 100 features [19]. Therefore, when the number of features remaining after the relevence and redundancy filters exceeds 110, we switch to the slightly less effective but more efficient Sequential Floating Forward Selection (SFFS) algorithm [20] We use a Mahalanobis distance measure inside both SFFS and SFBS. The ....

H.T. Ng, W.B. Goh and K.L. Low, "Feature Selection, Perceptron Learning, and a Usability Case Study for text Categorization ". 20th annual international ACM SIGIR conference on Research and development in information retrieval, July 27-31, Philadelphia, pp. 67-73, 1997.

....may be dubbed categorisation by context, since it exploits the context surrounding a link in an HTML document to extract useful information for categorising the document it refer to. This technique is complementary to the traditional technique of categorisation by content [Yang 94, Schtze 95, Ng 97] where information for categorising a document is extracted from the document itself. Such approach may exploit linguistic analysis to determine relevant portions of the text [Fuhr 91] and then exploits probabilistic or statistical analysis to perform feature selection [Yang 97] and to weight ....

....94, Schtze 95, Ng 97] where information for categorising a document is extracted from the document itself. Such approach may exploit linguistic analysis to determine relevant portions of the text [Fuhr 91] and then exploits probabilistic or statistical analysis to perform feature selection [Yang 97] and to weight selected features. Categorisation by context instead Attardi G. Di Marco S. Salvi D. Categorisation by Context 721 exploits relevance hints that are directly provided in the structure of the HTML documents which people build on the Web. Combining a large number of such hints, a ....

Ng, H.T., Goh, W.B., Low, K.L.: "Feature selection, perceptron learning, and a usability case study for text categorization"; Proceedings of SIGIR-97, 20th ACM International Conference on Research and Development in Information Retrieval, Philadelphia, USA (1997), 67--73.

....from several sources (general classification indexes, specialized thesauri, etc. Techniques for automatically deriving representations of categories ( category profile extraction ) and performing classification have been developed within the area of text categorization [Ittner 95, Lewis 96, Ng 97, Schtze 95, Yang 94, Yang 97] a discipline at the crossroads between information retrieval and machine learning. Text categorization uses machine learning techniques to inductively build representations of a given set of categories from a training set of documents pre categorized under them. An ....

....classification indexes, specialized thesauri, etc. Techniques for automatically deriving representations of categories ( category profile extraction ) and performing classification have been developed within the area of text categorization [Ittner 95, Lewis 96, Ng 97, Schtze 95, Yang 94, Yang 97] a discipline at the crossroads between information retrieval and machine learning. Text categorization uses machine learning techniques to inductively build representations of a given set of categories from a training set of documents pre categorized under them. An automatic process can then ....

Ng, H.T., Goh, W.B., Low, K.L.: "Feature selection, perceptron learning, and a usability case study for text categorization", Proceedings of SIGIR-97, 20th ACM International Conference on Research and Development in Information Retrieval, Philadelphia, US, 67--73, 1997.

....the majority of supervised IR for a long time [13] a number of recent studies have shown that alternative, more complex algorithms often outperform NB in terms of classification performance. Such algorithms originate from a variety of research areas such as Nearest Neighbor [14] neural networks [18], regression [19] rule induction [1] 6] and Support Vector Machines [11] The accuracy improvements are often significant in a variety of real world data sets (see also [20] for a comparative study of many algorithms) Support Vector Machines (SVM) have been recognized as one of the most ....

T.H.Ng, W.B.Goh, and K.L. Low, "Feature selection, perceptron learning and a usability case study for text categorization", 20 th ACM SIGIR Conference, 1997.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC