M. A. Hearst. Noun homograph disambiguation using local context in large corpora. In Proceedings of the 7th Annual Conference of the University of Waterloo Centre for the New Oxford English Dictionary, pages 1--22, Oxford, UK, 1991.  Home/Search   Document Details and Download   Summary   Related Articles   Check

.... dig is being used as a noun rather than as a verb indicates the word s appropriate meaning. But many words have multiple meanings even while occupying the same part of speech. To this end, the tagger has been used in the implementation of an experimental noun homograph disambiguation algorithm [Hearst, 1991]. The algorithm (known as Catch Word) performs supervised training over a large text corpus, gathering lexical, orthographic, and simple syntactic evidence for each sense of the ambiguous noun. After a period of training, CatchWord classifies new instances of the noun by checking its context ....

M. A. Hearst. Noun homograph disam- biguation using local context in large text corpora. In The Proceedings of the 7th New OED Conference on Using Corpora, pages 1-22, Oxford, 1991.

.... The references have ranged from detailed custom built lexicons (e.g. l 1] to standard resources like dictionaries and thesauri like Roget s (e.g. 2, 10, 14] To take the context into account, researchers have used a variety of statistical weighting and spreading activation models (e.g. [9, 14, 15]) This section gives brief descriptions of some approaches that use on line dictionaries and WordNet as references. WordNet is a large, manually constructed semantic network built at Princeton University by George Miller and his colleagues [12] The basic unit of WordNet is a set of synonyms, ....

Hearst, Marti. 1991. "Noun'Homograph Disambiguation Using Local Context in Large Text Corpora," Proceedings of the 7th Annual Conference of the UW Centre for the New OED and Text Research, Oxford, England.

....and selecttonal restrictions, and words occurring within a specified distance before and or after the target. An obvious shortcoming of this approach is the amount of work involved. Recently there has been much interest in automatic and semi automatic acquisition of local context (Hearst [2], Resnik [8] Yarowsky [13] These systems are all plagued with the same. problem, excellent precision but low recall. That is, if the local information that the methods learn is alo present in a novel context, then that information is very reliable. However, quite frequently no local context ....

Marti A. Hearst. Noun homograph disambiguation using local context in large text corpora. In Seventh Annual Conference of the I]W Centre for the IVew OED and Tezt Research: Using Corpora, pages 1-22, Oxford, 1991. UW Centre for the New OED and Text Research. 19

....called the circularity problem . Dagan and Itai, 1994) conclude that this is the most critical issue in developing feasible sense disambiguation methods. A few approaches to solve this problem are outlined below. 7.2. 1 Bootstrapping This approach construct the training corpus incrementally (Hearst, 1991). First, an initial set of occurrences for each sense is tagged manually. An initial model, M 1 , will be trained using this small corpus, C 1 . M 1 will then be used to disambiguate the rest of the occurrences of ambiguous words in the training corpus. All occurrences that can be disambiguated ....

....the occurrences of ambiguous words in the training corpus. All occurrences that can be disambiguated with strong con dence will be combine with C 1 to form C 2 . M 2 is then trained using C 2 . These steps are repeated until all the occurrences of the ambiguous words are tagged. The experiment of (Hearst, 1991) shows that to achieve high precision in word sense tagging, the initial set must be large (20 30 occurrences for each sense) Hence the cost of manually tagging the initial set is still high. 7.2.2 Clustering Occurrences of an Ambiguous Word This is an ecient mean of manual tagging proposed in ....

Hearst, M. (1991). Noun homograph disambiguation using local context in large text corpora. pages 1-22.

....(initially a small number of occurrences are manually tagged and used for training, then the system tags other occurrences which it is particularly confident about and these are also used for training, etc. have not so far solved the problem. Dagan and Itai s [4] opinion on Hearst s work [9]. 3.2 Comparison of systems described here Yarowsky s [1] thesaurus based system (or improved versions of similar) appears to be suitable for general purpose word sense disambiguation and achieves a high level of reliability without requiring tagged training data. Dagan and Itai s [4] system ....

Marti Hearst (1991), "Noun homograph disambiguation using local context in large text corpora", Proceedings of the Annual Conference of the UW Center for the New OED and Text Research pp. 1-22 Note: Hypertext links to most of the papers referenced are available on the Web at http://www2.eng.cam.ac.uk/~sm277/work/.

....each extension comes with a cost, in terms of escalating numbers of features and correspondingly sparse data. It is only viable to extend the repertoire of features if one also introduces methods for determining which are salient for each word. Papers exploring this route in different ways are (Hearst, 1991; Leacock, Towell, and Vorhees, 1993; Yarowsky, 1995; Pedersen, Bruce, and Wiebe, 1997) Note that if one sees the lexicon generation phase of c WSD as a one off, resource development activity, it becomes viable to spend substantially longer on it than if it is seen as a regularly repeated ....

Hearst, Marti A. 1991. Noun homograph disambiguation using local context in large text corpora.

.... (Salton and McGill, 1983; Croft, 1984; Turtle and Croft, 1992; Boostein, 1983; Korfhage, 1995) and sense disambiguation between multiple usages of the same word (Dagan and Itai, 1994; Gale et al. 1992a; Gale et al. 1992b; Gale et al. 1992c; Shutze, 1992; Gale et al. 1993; Yarowsky, 1995; Hearst, 1991). All these works are based on using content or context information as discriminatory features. In this section, we focus on the discussion of discriminant analysis using non parallel corpora. Dagan, 1990) was the first to use a pair of non parallel texts for the task of lexical disambiguation ....

M. Hearst. 1991. Noun homograph disambiguation using local context in large text corpora. In Using Corpora, Waterloo, Canada.

....trained on a New York Times News Service corpus. Our algorithm exceeds this accuracy on each word, with an average relative performance of 97 vs. 92 . 11 9 Comparison with Previous Work This algorithm exhibits a fundamental advantage over supervised learning algorithms (including Black (1988) Hearst (1991), Gale et al. 1992) Yarowsky (1993, 1994) Leacock et al. 1993) Bruce and Wiebe (1994) and Lehman (1994) as it does not require costly hand tagged training sets. It thrives on raw, unannotated monolingual corpora the more the merrier. Although there is some hope from using aligned ....

....that leveraging bilingual lexicons and monolingual 11 This difference is even more striking given that Schutze s data exhibit a higher baseline probability (65 vs. 55 ) for these words, and hence constitute an easier task. language models can overcome the need for aligned bilingual corpora. Hearst (1991) proposed an early application of bootstrapping to augment training sets for a supervised sense tagger. She trained her fully supervised algorithm on hand labelled sentences, applied the result to new data and added the most confidently tagged examples to the training set. Regrettably, this ....

Hearst, Marti, "Noun Homograph Disambiguation Using Local Context in Large Text Corpora," in Using Corpora, University of Waterloo, Waterloo, Ontario, 1991.

....achieved by judge 1, the algorithm, and random selection were respectively 68.6 , 60.5 , and 33.3 . As the relatively low accuracies for human judges demonstrate, disambiguation using WordNet s fine grained senses is quite a bit more difficult than disambiguation to the level of homographs (Hearst, 1991; Cowie, Guthrie, Guthrie, 1992) Resnik and Yarowsky (1997) discuss the implications of WordNet s fine grainedness for evaluation of word sense disambiguation, and consider alternative evaluation methods. Information Based Semantic Similarity However, many of the definitions do have a useful ....

Hearst, M. (1991). Noun homograph disambiguation using local context in large corpora.

.... dig is being used as a noun rather than as a verb indicates the word s appropriate meaning. But many words have multiple meanings even while occupying the same part of speech. To this end, the tagger has been used in the implementation of an experimental noun homograph disambiguation algorithm [ Hearst, 1991 ] The algorithm (known as CatchWord) performs supervised training over a large text corpus, gathering lexical, orthographic, and simple syntactic evidence for each sense of the ambiguous noun. After a period of training, CatchWord classifies new instances of the noun by checking its context ....

M. A. Hearst. Noun homograph disambiguation using local context in large text corpora. In The Proceedings of the 7th New OED Conference on Using Corpora, pages 1--22, Oxford, 1991.

.... of a monolingual text and context information of its words have been used in algorithms for author characterization from documents [26] Lecture Notes in Computer Science 7 document categorization from queries [1, 5, 24, 27, 30] and sense disambiguation between multiple usages of the same word [8, 20, 19, 17, 18, 22, 28, 33]. We propose to use context information of a word to find its counter part in the other language in a comparable corpus. Our goal is to find translation or translation candidates for new words which are not found in an online dictionary, and then use the result to augment the dictionary, in order ....

M. Hearst. Noun homograph disambiguation using local context in large text corpora. In Using Corpora, Waterloo, Canada, 1991.

....by the top ranked definition was chosen as the sense of w. Since Lesk s paper a bewildering array of disambiguators have been built: Cowie [6] Black [7] Wallis [8] and Demetriou [9] have made further use of dictionaries; Zernik [10] built a disambiguator using a morphological analyser; Hearst [11] used learning based on human evidence; Dagan [12] used bilingual corpora; Church [13] tried aligned bilingual corpora; Voorhees [14] and Sussna [15] used the WordNet thesaurus; and Yarowsky [16] used a combinationofRoget sthesaurusand Grollier s encyclopaedia to produceoneofthe better ....

Hearst MA. Noun homograph disambiguation using local context in large text corpora. Proceedings of the 7th conference, UW Centre for the New OED & Text Research Using Corpora, 1991; 7

....you might be able to do with these exciting new resources. Since then, with the advent of language corpora and the rapid growth of statistical work in NLP, the number of possibilities for how you might go about WSD has mushroomed, as has the quantity of work on the subject (Brown et al. 1991; Hearst, 1991; McRoy, 1992; Gale, Church, and Yarowsky, 1992; Yarowsky, 1992) Clear, 1994) Sch utze and Pederson, 1995) and (Yarowsky, 1995) are of particular interest because of their approach to the issue of the set of word senses to be disambiguated between. Sch utze and Pederson devised ....

Hearst, Marti A. 1991. Noun homograph disambiguation using local context in large text corpora. In Using Corpora: Proc. Seventh Ann. Conf. of the UW Centre for the New OED, pages 1--22, Waterloo, Canada.

....was observed in the training data. Finding a more accurate probability estimate depends on several factors, including the size 4 The richness of this feature set is one of the key reasons for the success of this algorithm. Others who have very productively exploited a diverse feature set include [Hea91], Bri93] and [DI94] 5 Position markers include 1 (token to the right) Gamma1 (token to the left) Sigmak (co occurrence in Sigmak token window) and V (head verb) Possible types of objects at these positions include w (raw words) p (parts of speech) and l (lemmas a class of words ....

M. Hearst. Noun homograph disambiguation using local context in large text corpora. In Using Corpora, University of Waterloo, Waterloo, Ontario, 1991.

....corpora are not available. Warren (1978) gives frequency counts on manually analyzed, coarse semantic relation categories in her study of nominal compounds, but these must be massaged to fit more realistic AI ontologies. Another potential use of large corpora techniques has been suggested by Hearst (1991), who proposes an automated method for coarse disambiguation of noun homographs that could be used as a front end hypothesis generator. Such a method, based on orthographic, lexical, and syntactic cues near the noun, may improve the relevance and accuracy rate of hypotheses. Performance will ....

HEARST, MARTI A. 1991. Noun homograph disambiguation using local context in large text corpora. In Seventh Annual Conference of the University of Waterloo Centre for the New OED and Text Research: Using Corpora, 1--22, Oxford.

.... Others have also argued that the task of simplifying lexical entries on the basis of broad semantic class membership is complex and, perhaps, infeasible (see, e.g. Boguraev and Briscoe (1989) However, a number of researchers (Fillmore, 1968; Grimshaw, 1990; Gruber, 1965; Guthrie et al. 1991; Hearst, 1991; Jackendoff, 1983; Jackendoff, 1990; Levin, 1993; Pinker, 1989; Yarowsky, 1992) have demonstrated conclusively that there is a clear relationship between syntactic context and word senses; it is our aim to exploit this relationship for the acquisition of semantic lexicons. We first describe the ....

Hearst, M. 1991. Noun Homograph Disambiguation Using Local Context in Large Text Corpora.

....326 600 Figure 4: interest EM Feature Set A (left) and D (right) 6 Related Work Bootstrapping approaches to word sense disambiguation require a small amount of disambiguated text in order to initialize the unsupervised learning algorithm. An early example of such an approach is described in [12]. A supervised learning algorithm is trained with a small amount of manually sense tagged text and applied to a held out test set. Those examples in the test set that are most confidently disambiguated are added to the training sample. A more recent bootstrapping approach is described in [37] ....

M. Hearst. Noun homograph disambiguation using local context in large text corpora. In Proceedings of the 7th Annual Conference of the UW Centre for the New OED and Text Research: Using Corpora, Oxford, 1991.

....approaches that rely only on the features in a text that can be automatically identified. 7.1 Bootstrapping Bootstrapping approaches require a small amount of disambiguated text in order to initialize the unsupervised learning algorithm. An early example of such an approach is described in (Hearst, 1991). A supervised learning algorithm is trained with a small amount of manually sense tagged text and applied to a held out test set. Those examples in the test set that are most confidently disambiguated are added to the training sample. A more recent bootstrapping approach is described in ....

Hearst, M. 1991. Noun homograph disambiguation using local context in large text corpora. In Proceedings of the 7th Annual Conference of the UW Centre for the New OED and Text Research: Using Corpora, Oxford.

No context found.

M. A. Hearst. Noun homograph disambiguation using local context in large corpora. In Proceedings of the 7th Annual Conference of the University of Waterloo Centre for the New Oxford English Dictionary, pages 1--22, Oxford, UK, 1991.

No context found.

Marti A. Hearst. Noun Homograph Disambiguation Using Local Context in Large Text Corpora. In Proceedings of the 7th Annual Conference of the University of Waterloo Centre for the the New OED and Text Research: Using Corpora, Oxford, 1991.

No context found.

Hearst, M. A. (1991). Noun homograph disambiguation using local context in large text corpora. In Using Corpora, U. of Waterloo. Hripcsak, G., C. Friedman, P. O. Alderson, W. DuMouchel, S. B.

No context found.

M. A. Hearst. Noun homograph disambiguation using local context in large corpora. In Proceedings of the 7th Annual Conference of the University of Waterloo Center for the New Oxford English Dictionary, pages 1--22, Oxford, UK, 1991.

No context found.

Hearst, Marti, "Noun Homograph Disambiguation Using Local Context in Large Text Corpora," in Us- ing Corpora, University of Waterloo, Waterloo, On- tario, 1991.

No context found.

M. Hearst. Noun homograph disambiguation using local context in large text corpora. In Using Corpora, University of Waterloo, Waterloo, Ontario, 1991.

No context found.

Marti A. Hearst. 1991. Noun Homograph Disambiguation Using Local Context in Large Text Corpora. In Using Corpora, University of Waterloo, Waterloo, Ontario.

No context found.

Hearst, Marti 1991. Noun Homograph Disambiguation Using Local Context in Large Text Corpora. In Proceedings of the 7th Annual Conference of the UW Centre for the New oed and Text Research, Oxford, England.

No context found.

Hearst, Marti, "Noun Homograph Disambiguation Using Local Context in Large Text Corpora," in Using Corpora, University of Waterloo, Waterloo, Ontario, 1991.

No context found.

Hearst, M., "Noun Homograph Disambiguation Using Local Context in Large Text Corpora," Using Corpora, University of Waterloo, Waterloo, Ontario, 1991.

No context found.

M. Hearst. Noun homograph disambiguation using local context in large text corpora. In Using Corpora, University of Waterloo, Waterloo, Ontario, 1991.

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