Z. Wu and M. Palmer. Verbs semantics and lexical selection. In Proceedings of the 32nd annual meeting on Association for Computational Linguistics, pages 133--138, Morristown, NJ, USA, 1994. Association for Computational Linguistics.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....keywords and enhances hyperlinks with semantics, by mapping sets of keywords that describe a web page, to sets of concepts (categories) organized in a hierarchy. In the current implementation the mechanism employs WordNet [WN] as a thesams and ontology, the Wu and Palmer similarity measure [WP94] for computing similarities between terms in a hierarchy. a novel similarity measure for weighted sets of terms in a hierarchy, that lets us use any sort of proximity based clustering algorithm (such as DBSCAN [EK 96] a prototype client server system, called THESUS, that a. collects URLs and ....

....similarity measures in the more simple case of calculating the similarity between two given terms of the ontology. RSM94] and [Res95, Res99] propose different measure inside a taxonomy such as WordNet, and [Lin98] proposes a comparison between these measures and others, such as Wu and Palmer [WP94], Miller and Charles, and a novel similarity measure. DJ02] also propose the use of Wu and Palmer measure in the ontology context. The Wu and Palmer measure is the fastest to compute, and is arguably as good as the others (see [Lin98] which is why we chose to use it. We detail its definition ....

[Article contains additional citation context not shown here]

Z. Wu and M. Palmer "Verb Semantics and Lexical Selection", Proceedings of the 32nd Annual Meetings of the Associations for Computational Linguistics, pages 133-138.

....in MT systems that are organized by language specific entries. In an earlier description of this work, we identified the IL lexicon as a concept based lexicon. Dorr et al. 1994] 2SThe nodes in our figures do not have their LOC field marked. 29While this approach resembles that taken by [Wu and Palmer, 1994], who address related issues in lexical selection, our IL lexicon is structured by the syntax of our IL while their hierarchy is defined in terms of concepts that encode a multi part meaning representation. GO LOC( Y, ATH UP E lift (Z) Ti kaldk [CAUSE (KY) j . GO LOC(Y, PATll ....

Z. Wu and M. Palmer. "Verb Se- mantics and Lexical Selection". In Proceedings of the $2nd Annual Meeting of the Association for Computational Linguistics, pages 133 138, Las Cruces, New Mexico, 1994.

....c G 1 p and c G 2 p have different types, i.e. refer to elements of different classes. In this case, we define b ( c G 1 p , c G 2 p ) as the semantic similarity between type( c G 1 p ) and type( c G 2 p ) in the is a hierarchy. We calculate it using a similar expression to one proposed in [11]. In this third option of our formula, d i indicates the distance number of nodes from the type i to the root of the hierarchy. The relational similarity s r expresses how similar the relations among the common concepts in the conceptual graphs G 1 and G 2 are. In other words, the ....

Wu and Palmer (1994), "Verb Semantics and Lexical Selection", Proc. of the 32nd Annual Meeting of the Associations for Computational Linguistics, 1994.

....concepts. We first define the similarity measure between two concepts which makes it possible to evaluate the semantic distance between these two concepts. This measure is defined relatively to a thesaurus and to the hypernyms relationship. In our context, we use the similarity measure defined by [29]. They propose a similarity measure related to the edge distance in the way it takes into account the most specific subsumer of the two concepts, characterizing their commonalities, while normalizing in a way that accounts for their differences. Their measure is shown in formula 3 where , is the ....

....relationship. So, all pages containing these concepts will be rejected. The near operator is not related to the distance between words (number of words between two words) as in the classical approach. But, it is related to a semantic distance between concepts according to the similarity measure [29] used to calculate the representativeness coefficient. In our context, the near operator becomes an unary operator and makes it possible to add to the query all the concepts semantically connected to the targeted concept and in its neighbourhood. 7 Related works Our choices differ from related ....

Z. Wu and M. Palmer, "verb semantics and lexical selection", In Proceedings of the 32nd annual meeting of the association for computational linguistics, Las Cruses, New Mexico, 1994

.... given that, how do we come to understand the entire utterance, including any inferences that the utterance carries with it These questions correspond to well defined issues in natural language processing, specifically in Machine Translation: the problems of sense disambiguation and lexical choice [Wu and Palmer, 1995]. In the machine translation domain, the process of understanding which sense of a verb like open is used in the source language is a sense disambiguation task; the source utterance must be interpreted before beginning to translate. And the problem of selecting the best word (or set of words) to ....

Zhibiao Wu and Martha Palmer. Verb semantics and lexical selection. In Proceedings of the 32nd Annual Meeting of the Association for Computational Linguistics, 1995.

....of the distances between the pairs of adjacent nodes along the shortest path connecting c i and c j . 2.3.3. 2 Wu and Palmer s Conceptual Similarity In a paper focusing on semantic representation of verbs in computer systems and its impact on lexical selection problems in machine translation , Wu and Palmer [1994] devote a couple of paragraphs to introducing a metric that is somewhat specialized but nonetheless deserving of at least a brief mention. Very superficially, the key idea of the authors approach to translating English verbs into Mandarin Chinese is to project verbs (and verb compounds) of both ....

....incorporated therein) Jabber s efficacy rests primarily on the success of lexical chaining as an information retrieval mechanism . 4.4. Lexical Selection 67 4.4 Lexical Selection 4.4.1 Wu and Palmer Having defined the similarity measure in a conceptual domain (Equation 2.11, Section 2.3.3. 2) Wu and Palmer [1994] propose to define similarity between two verb meanings, e.g. a target verb and a source verb . as a summation of weighted similarities between pairs of simpler concepts in each of the domains the two verbs are projected onto . Formally, simWP (v 1 ; v 2 ) X i W i Theta simWP (c i;1 ; ....

Zhibiao Wu and Martha Palmer. Verb semantics and lexical selection. In Proceedings of the 32nd Annual Meeting of the Association for Computational Linguistics, pages 133--138, Las Cruces, New Mexico, June 1994.

....minimality, although intuitively appealing, may not be a necessary property of a similarity measure faithful to human judgments; since Lin s measure ensures minimality and the information content measure does not, further experimentation is clearly needed to evaluate their competing predictions. Wu and Palmer (1994) propose a similarity measure that is based on edge distances, but related to Lin s measure in the way it takes into account the most specific node dominating c 1 and c 2 , characterizing their commonalities, while normalizing in a way that accounts for Information Based Semantic Similarity ....

Wu, Z., & Palmer, M. (1994). Verb Semantics and Lexical Selection. In Proceedings of the 32nd Annual Meeting of the Association for Computational Linguistics Las Cruces, New Mexico.

....similarity measures do not correlate to human judgments as well as his measure. Resnik s similarity measure is quite close to the one proposed here: sim Resnik (A; B) 1 2 I(common(A; B) For example, in Figure 2, sim Resnik (Hill; Coast) Gamma log P (Geological Formation) Wu and Palmer [Wu and Palmer, 1994] proposed a measure for semantic similarity that could be regarded as a special case of sim(A; B) simWu Palmer (A; B) 2 Theta N 3 N 1 N 2 2 Theta N 3 where N 1 and N 2 are the number of IS A links from A and B to their most specific common superclass C; N 3 is the number of IS A links ....

Wu, Z. and Palmer, M. (1994). Verb semantics and lexical selection. In Proceedings of the 32nd Annual Meeting of the Associations for Computational Linguistics, pages 133--138, Las Cruces, New Mexico.

.... SEMANTICS AND LEXICAL SELECTION ZHIBIAO WU NATIONAL UNIVERSITY OF SINGAPORE 1994 VERB SEMANTICS AND LEXICAL SELECTION BY ZHIBIAO WU A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY NATIONAL UNIVERSITY OF SINGAPORE 1994 i ACKNOWLEDGEMENTS I wish to express my deep gratitude to my supervisor Professor Hsu Loke Soo, for his stimulating insights, constant ....

Wu, Z. & Palmer, M. (1994). Verb Semantics and Lexical Selection. In Proceedings of ACL94, New Mexico State University, USA.

.... semantic distinctions [5, 7] or even require reference to the context [3] For example, the semantic features that are used in selecting the correct serial verb construction in Chinese (such as the initial shape of the object, choice of instrument) are not all used in selecting English verb senses [21]. The end result is that lexical selection is often predicated on the existence of semantic features that are completely irrelevant to the source language 2 . In Korean, there are several different senses of WEAR that use the same lexical item in English but require distinct Korean ....

Zhibiao Wu and Martha Palmer. Verb semantics and lexical selection. In 32nd Meeting of the Association for Computational Linguistics, Las Cruces, New Mexico, 1994.

....no way of gracefully adapting to the best partial match. The importance of allowing for best partial matches, in particular for translation purposes, has been demonstrated by an experimental system for English Chinese machine translation system implemented at the National University of Singapore [20], 15] This system achieved a rate of successful lexical choice that was significantly higher than the TRANSTAR [15] accuracy rate for the break examples from the Brown corpus, as well as additional sentences featuring other break verbs and hit, touch and cut verbs. Note that accurate lexical ....

Zhibiao Wu and Martha Palmer. Verb semantics and lexical selection. In Proceedings of the 32nd Annual Meeting of the Association for Computational Linguistics, New Mexico State University, July 1994.

....necessary to further filter the choice of the lexical item. 1 Introduction One of the primary tasks for machine translation is lexical selection selecting the target lexical item that most closely matches the source lexical item being translated. For transfer based approaches such as Transtar [17] and Geta [15] each separate lexeme in the source language must be paired with a corresponding lexeme in the target language in a set of bilingual dictionaries. An alternative 1 is the interlingua approach, such as Princitran [3] or Translator [8] in which the source verb is mapped to a ....

....system such as TRANSTAR, a commercial broad coverage English Chinese MT system developed in Beijing. When this system is applied to sentences from the Brown corpus that contain break, an accuracy rate of less than 20 is achieved, even after ruling out idiomatic uses and problems with parsing [17]. The primary reason is that in English break can be thought of as a very general verb indicating an entire set of breaking events which can be distinguished by the resulting state of the object being broken. Shatter, snap, split, etc. can all be seen as more specialized versions of the general ....

Zhibiao Wu and Martha Palmer. Verb semantics and lexical selection. In 32nd Meeting of the Association for Computational Linguistics, Las Cruces, New Mexico, 1994.

.... or metaphorical interpretation, but also an extension of meaning to a broader class of arguments, as in the extension of break from broken wire, meaning separated into pieces, into broken insulation, meaning a separation of the surface, 18] or from break the fence to break the language barrier [23]. We will illustrate the difficulty of this task with examples involving the translation of English break to Chinese. We propose that the representation of each sense of an individual lexical item must include the ways in which it is related to other similar senses which semantic concepts are ....

....usages. 2. Lexical semantic divergences After close examination of appropriate translations of English break expressions into Chinese (Mandarin) we have determined that English and Chinese are quite far apart in their representation of breaking events, as in John broke the window with a hammer, [23]. There are several factors that contribute to this divergence. The most significant difference is that Chinese uses a compound Verb Adjective construction that makes both the action precipitating the change of state and the 4 details of the resulting state explicit. Although English also makes ....

Zhibiao Wu and Martha Palmer. Verb semantics and lexical selection. In Proceedings of ACL94, New Mexico State University, USA, May 1994.

No context found.

Z. Wu and M. Palmer. Verbs semantics and lexical selection. In Proceedings of the 32nd annual meeting on Association for Computational Linguistics, pages 133--138, Morristown, NJ, USA, 1994. Association for Computational Linguistics.

No context found.

Wu, Z. and M. Palmer (1994). Verb semantics and lexical selection. 32nd Annual Meeting of the Association for Computational Linguistics. Conference, San Francisco, Morgan Kaufmann Publishers.

No context found.

Z. Wu, M. Palmer. Verb semantics and lexical selection, In 32nd Annual Meeting of the Association for Computational Linguistics, Las Cruces, New Mexico, pp. 133--138, 1994.

No context found.

Z. Wu and M. Palmer. Verb Semantics and Lexical Selection. In 32nd. Annual Meeting of the Association for Computational Linguistics, pages 133--138, New Mexico State University, Las Cruces, New Mexico, 1994.

No context found.

Z. Wu and M. Palmer. 1994. Verb semantics and lexical selection. In Proceedings of the 32nd Annual Meeting of the Association for Computational Linguistics, Las Cruces, New Mexico.

No context found.

Z. Wu and M. Palmer, `Verb semantics and lexical selection', in Proceedings of the 32nd Annual Meeting of the Association for Computational Linguistics (ACL-94), Las Cruces, NM, 133-138, (1994).

No context found.

Z. Wu, M. Palmer, Verb semantics and lexical selection, in: 32nd Annual Meeting of the Association for Computational Linguistics, Las Cruces, New Mexico, 1994, pp. 133--138.

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