Y. Karov and S. Edelman. Learning similarity-based word sense disambiguation from sparse data. In E. Ejerhed and I. Dagan, editors, Proceedings of the Fourth Workshop on Very Large Corpora, Copenhagen, 1996.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....systems now incorporate both statistical and symbolic linguistic knowledge; this trend led Gazdar to write of paradigm merger in [Gazdar 96] Secondly, research into statistical NLP has become more interested in models . that yield good results with relatively small samples [Dunning 93, Karov Edelman 96] Hybrid systems are indeed increasingly common, but it may be too early to talk of consensus breaking out into full paradigm merger. Charniak wrote in 1993 that . it is fair to say that few, if any, consider the traditional study of language from an artificial intelligence point of view a ....

Y. Karov and S. Edelman. Learning similarity-based word sense disambiguation from sparse data. In E. Ejerhed and I. Dagan, editors, Proceedings of the Fourth Workshop on Very Large Corpora, Copenhagen, 1996.

....systems now incorporate both statistical and symbolic linguistic knowledge; this trend led Gazdar to write of paradigm merger in [Gazdar 96] Secondly, research into statistical NLP has become more interested in models . that yield good results with relatively small samples [Dunning 93, Karov Edelman 96] Hybrid systems are indeed increasingly common, but it may be too early to talk of consensus breaking out into full paradigm merger. Charniak wrote in 1993 that . it is fair to say that few, if any, consider the traditional study of language from an artificial intelligence point of view a ....

Y. Karov and S. Edelman. Learning similarity-based word sense disambiguation from sparse data. In E. Ejerhed and I. Dagan, editors, Proceedings of the Fourth Workshop on Very Large Corpora, Copenhagen, 1996.

....learned for translation lexicons are string to string mappings, instead of the string to structure mappings learned by Wolfie. 7.2. 2 Acquisition from MRDs Many researchers (Amsler, 1981; Walker Amsler, 1986; Byrd, Calzolari, Chodorow, Klavans, Neff, Risk, 1987; Boguraev Briscoe, 1989; Karov Edelman, 1996; Rigau, Rodr iguez, Agirre, 1998) have investigated the extraction of lexical information from Machine Readable Dictionaries (MRDs) While these methods may be capable of generating general information, additional information is needed from some source to tailor the information to each ....

Karov, Y., & Edelman, S. (1996). Learning similarity-based word sense disambiguation from sparse data. In Proceedings of the Fourth Workshop on Very Large Corpora Copenhagen.

....According to this formula, w 2 is more likely to occur with w 1 if it tends to occur with the words that are most similar to w 1 . Considerable latitude is allowed in defining the set S(w 1 ) as is evidenced by previous work that can be put in the above form. Essen and Steinbiss (1992) and Karov and Edelman (1996) (implicitly) set S(w 1 ) V 1 . However, it may be desirable to restrict S(w 1 ) in some fashion, especially if V 1 is large. For instance, Dagan, Pereira, and Lee (1994) use the closest k or fewer words w 0 1 such that the dissimilarity between w 1 and w 0 1 is less than a threshold value ....

Karov, Yael and Shimon Edelman. 1996. Learning similarity-based word sense disambiguation from sparse data. In 4rth Workshop on Very Large Corpora.

....on the similarity between x and x 0 . We are not the originators of equation (4. 1) Similarity based estimation was first used for language modeling in the cooccurrence smoothing method of Essen and Steinbiss (1992) derived from work on acoustic model smoothing by Sugawara et al. 1985) Karov and Edelman (1996) develop a similarity based disambiguation method that also can be fit into the framework of equation (4.1) however, since their method does not estimate probabilities and relies on a similarity function that is calculated via an iterative process, we will not give further consideration to their ....

....x 0 ) P x 0 2S(x) W (x; x 0 ) P (yjx 0 ) 4.2) Observe that according to this formula, we predict that y is likely to occur with x if it tends to occur with objects that are very similar to x. Considerable latitude is allowed in defining the set S(x) Essen and Steinbiss (1992) and Karov and Edelman (1996) (implicitly) set S(x) X . However, if X is very large, it is desirable to restrict S(x) in some fashion, so that summing over all x 0 2 X is not too time consuming. In the next chapter, we will consider various heuristics for choosing a small set of similar words. These heuristics include ....

Karov, Yael and Shimon Edelman. 1996. Learning similarity-based word sense disambiguation from sparse data. In 4rth Workshop on Very Large Corpora. Also available as CS-TR 96-05, The Weizmann Institute of Science.

....precision 71.2 and recall 61.4 for nouns in four randomly selected semantic concordance files. From among the methods based on semantic distance, Resnik, 1993) Sussna, 1993) use a similar semantic distance measure for two concepts in WordNet, but they also focus on selected group of nouns only. (Karov and Edelman, 1996) use an interesting iterative algorithm and attempt to solve the sparse data bottleneck by using a graded measure of contextual similarity. They achieve 90.5, 92.5, 94.8 and 92.3 percent accuracy in distinguishing between two senses of the noun drug, sentence, suit and player, respectively. ....

Y. Karov and S. Edelman. 1996. Learning similarity-based word sense disambiguation from sparse data. In Proc. of the 3rd Workshop on Very Large Corpora, pages 42--55.

....According to this formula, w 2 is more likely to occur with w 1 if it tends to occur with the words that are most similar to w 1 . Considerable latitude is allowed in defining the set S(w 1 ) as is evidenced by previous work that can be put in the above form. Essen and Steinbiss (1992) and Karov and Edelman (1996) (implicitly) set S(w 1 ) V 1 . However, it may be desirable to restrict S(w 1 ) in some fashion for efficiency reasons, especially if V 1 is large. For instance, in the language modeling application of Section 3, we use the closest k or fewer words w 0 1 such that the dissimilarity between w ....

....the nearest cluster which determines the sense for that occurrence. Schutze emphasizes that his method avoids clustering words into a pre defined set of classes, claiming that such clustering is likely to introduce artificial boundaries that cut off words from part of their semantic neighborhood. Karov and Edelman (1996) have also addressed the data sparseness problem in word sense disambiguation by using word similarity. They use a circular definition for both a word similarity measure and a context similarity measure. The circularity is resolved by an iterative process in which the system learns a set of ....

Karov, Y., & Edelman, S. (1996). Learning similarity-based word sense disambiguation from sparse data. In E. Ejerhed & I. Dagan (Eds.), Fourth Workshop on Very Large Corpora (pp. 42--55).

....is essential for NLP to attain high level natural language inference. Contrary to this intuition, experiments in text retrieval and natural language have not shown much improvement when incorporating information of the kind humans seem to use (Krovetz and Croft, 1992; Kosmynin and Davidson, 1996; Karov and Edelman, 1996; Junker, 1997) The lack of significant improvement in the presence of more knowledge may be explained by the type of knowledge used, the way it is incorporated, and the learning algorithms employed. In the present paper we study an effective way of incorporating incomplete and ambiguous ....

Y. Karov and S. Edelman. 1996. Learning similaritybased word sense disambiguation from sparse data. In Fourth workshop on very large corpora, pages 42--55, August.

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