C. Leacock and M. Chodorow. Combining local context and wordnet similarity for word sense identification. WordNet: An Electronic Lexical Database, 49(2):265--283, 1998.  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... to the qvar, the reciprocal of the length of the path between the type of the qvar entity and the type of eY in the semantic lattice (ontology) or if this fails (usually because the two entities are not both presem in the system s small ontology) the reciprocal of the Leacock Chodorow distance [10] between the qvar and eY in WordNet [12] For instance if qvar and eY are of the same type then they will receive a score of 1. b) Object Relatiom O: 0.25 if eY is related to a question constraint within the sentence by apposition, a qualifying relationship, or with the prepositions of or n. ....

C. Leacock and M. Chodorow. Combining Local Context and WordNet Similarity for Vord Sense Identification. In C. Fellbaum, editor, WordNet: An Electronic Lexical Database, chapter 11, pages 265 285. MIT Press, 1998.

....as provided by WordNet. Please note that in the rest of this paper concept and word sense are used somewhat interchangeably, since each concept in WordNet represents a distinct meaning that can be considered a word sense. 4. 1 The Leacock Chodorow Measure The measure of Leacock and Chodorow [8] is based on the lengths of paths between noun concepts in an is a hierarchy. The shortest path between two concepts is the one which includes the fewest number of intermediate concepts. This value is scaled by the depth of the hierarchy, where depth is de ned as the length of the longest path ....

....sense. Though this method does not specify an exact formula for semantic relatedness of words, it appears to be built upon node counting techniques for measuring semantic relatedness and gives us yet another way to use the WordNet is a hierarchy for word sense disambiguation. Leacock and Chodorow [8] have used their measure to augment a supervised approach to word sense disambiguation that relies on local context, which are features that occur in close proximity to the target word. They use their measure (as well as Resnik s) to determine the relatedness between a noun in each test instance ....

C. Leacock and M. Chodorow. Combining local context and WordNet similarity for word sense identi cation. In C. Fellbaum, editor, WordNet: An electronic lexical database, pages 265-283. MIT Press, 1998.

....and the topology of the multimedia knowledge. There are many proposed methods for calculating semantic distance or similarity among concepts in semantic concept networks such as WordNet. Some methods rely uniquely on the hierarchical specialization generalization relationships among concepts [12,13] whereas others take into account all the semantic relations [19] There are methods that use exclusively the concept network topology [13,19] while others combine both concept network topology information and text corpus statistics (e.g. concept probabilities) 12] The most commonly used ....

....in semantic concept networks such as WordNet. Some methods rely uniquely on the hierarchical specialization generalization relationships among concepts [12,13] whereas others take into account all the semantic relations [19] There are methods that use exclusively the concept network topology [13,19] while others combine both concept network topology information and text corpus statistics (e.g. concept probabilities) 12] The most commonly used concept network for calculating semantic relatedness is WordNet [12,13,19] Recent work evaluated five semantic distance measures using WordNet [6] ....

[Article contains additional citation context not shown here]

Leacock, C., and M. Chodorow, "Combining Local Context and WordNet Similarity for Word Sense Identification", Fellbaum, pp. 265-283, 1998.

....in our experiments. Our application work is focused on the spatial domain so, our experiments employ subsets of the two readily available resources, WordNet (334 definitions) 9] and SDTS (498 definitions) 41] that deal with spatial concepts. WordNet is a widely used terminological ontology [4, 56 58] that organizes concepts in sets of synonyms (synsets) connected by semantic relations. It contains approximately 118,000 words organized into 90,000 sets of synonyms, which are semantically interrelated depending on their syntactic category. SDTS was created to provide a common classification and ....

Leacock, C. and M. Chodorow, Combining Local Context and WordNet Similarity for Word Sense Identification, in C. Fellbaum (ed.), WordNet: An Electronic Lexical Database 1998, The MIT Press: Cambridge, MA. p. 265-283.

....and its role may vary among different domains [23] For NLP, context has a sense disambiguation function [18] so that otherwise ambiguous statements become meaningful and precise. Studies in NLP analyze the meaning of words within either a topical context or a local context of a corpus [24]. Knowledge representation involves statements and axioms that hold in certain contexts; therefore, context determines the truth or falsity of a statement as well as its meaning [19] For knowledge based problem solving, context is usually defined as the situation or circumstances that surround a ....

Leacock, C. and M. Chodorow, 1998, Combining Local Context and WordNet Similarity for Word Sense Identification, in: C. Fellbaum (editor) WordNet: An Electronic Lexical Database. pp. 265-283, The MIT Press: Cambridge, MA.

....(######## # ###) Hyponym i.e. speci cation (##### # ######) and Hypernym i.e. generalization (##### # #####) 2.3 Semantic Similarity of Words Several measures of semantic similarity based on distance between words in WordNet have been used by di erent researchers. Leacock and Chodorow [13] have used the negative logarithm of the normalized shortest path length as a measure of similaritybetween twowords, where the path length is measured as the number of nodes in the path between the twowords and the normalizing factor is the maximum depth in the taxonomy. In this metric, the ....

....#(# # ## # ) across all pairs of words (# # ## # ) The noun hierarchy of the WordNet is disconnected there are 11 separate hierarchies with distinct root nodes. The verb hierarchy is also disconnected, with 15 distinct root nodes. For our purpose, following the method of Leacock and Chodorow [13], wehave connected the 11 root nodes of the noun hierarchy to a single root node ##### so that a path can always be found between two nouns. Similarly,we have connected the verb root nodes by a single root node ##### . ##### and ##### are further connected to a toplevel root node, #### . This ....

C. Leacock and M. Chodorow. Combining local context and WordNet similarity for word sense identication. In C. Fellbaum, editor, ######## ## ########## ####### ########,chapter 11, pages 265-284. MIT Press, 1998.

....(computer cpu) Hyponym i.e. speci cation (plant fungus) and Hypernym i.e. generalization (apple fruit) 2.3 Semantic Similarity of Words Several measures of semantic similarity based on distance between words in WordNet have been used by di erent researchers. Leacock and Chodorow [13] have used the negative logarithm of the normalized shortest path length as a measure of similarity between two words, where the path length is measured as the number of nodes in the path between the two words and the normalizing factor is the maximum depth in the taxonomy. In this metric, the ....

....i ; w j ) across all pairs of words (w i ; w j ) The noun hierarchy of the WordNet is disconnected there are 11 separate hierarchies with distinct root nodes. The verb hierarchy is also disconnected, with 15 distinct root nodes. For our purpose, following the method of Leacock and Chodorow [13], we have connected the 11 root nodes of the noun hierarchy to a single root node Rnoun so that a path can always be found between two nouns. Similarly, we have connected the verb root nodes by a single root node Rverb . Rnoun and Rverb are further connected to a toplevel root node, Rtop . This ....

C. Leacock and M. Chodorow. Combining local context and WordNet similarity for word sense identication. In C. Fellbaum, editor, WordNet: An Electronic Lexical Database, chapter 11, pages 265-284. MIT Press, 1998.

....words immediately to the right and left of the target word to be the context in which it appears. This narrow window is called the local context of the word [7] One could also look at other words with which the target word co occurs. This broader definition of context is known as topical context [5, 6]. Many approaches to word sense disambiguation use statistics gleaned from large amounts of text that are hand tagged with the correct answers [5, 6, 7] to analyze a new text. The handtagged text is known as training data. Though local context has been shown to be e#ective in disambiguating the ....

....word [7] One could also look at other words with which the target word co occurs. This broader definition of context is known as topical context [5, 6] Many approaches to word sense disambiguation use statistics gleaned from large amounts of text that are hand tagged with the correct answers [5, 6, 7] to analyze a new text. The handtagged text is known as training data. Though local context has been shown to be e#ective in disambiguating the senses of a particular word form [5] the use of topical context has the advantage that it makes maximal use of small sets of training data. Maximizing ....

[Article contains additional citation context not shown here]

Claudia Leacock and Martin Chodorow. Combining local context and wordnet similarity for word sense identification. In WordNet, an Electronic Lexical Database, pages 285--303. MIT Press, Cambridge MA, 1998.

....of plant. 11. Hypernym: This pointer refers to a generalization of the concept, e.g. fruit is a hypernym of apple. 2.3 Semantic Similarity of Words Several measures of semantic similarity based on distance between words in WordNet have been used by di erent researchers. Leacock and Chodorow [LC98] have used the negative logarithm of the normalized shortest path length as a measure of similarity between two words, where the path length is measured as the number of nodes in the path between the two words and the normalizing factor is the maximum depth in the taxonomy. In this metric, the ....

....value of d(w i ; w j ) across all pairs of words (w i ; w j ) The noun hierarchy of the WordNet is disconnected there are 11 trees with distinct root nodes. The verb hierarchy is also disconnected, with 15 distinct root nodes. For our purpose, following the method of Leacock and Chodorow [LC98] we have connected the 11 root nodes of the noun hierarchy to a single root node R noun so that a path can always be found between two nouns. Similarly, we have connected the verb root nodes by a single root node R verb . R noun and R verb are further connected to a top level root node, R top . ....

C. Leacock and M. Chodorow. Combining local context and WordNet similarity for word sense identication. In C. Fellbaum, editor, WordNet: An Electronic Lexical Database, chapter 11, pages 265-284. MIT Press, 1998.

No context found.

C. Leacock and M. Chodorow. Combining local context and wordnet similarity for word sense identification. WordNet: An Electronic Lexical Database, 49(2):265--283, 1998.

No context found.

C. Leacock, M. Chodorow. Combining local context and WordNet similarity for word sense identification, In Fellbaum MIT Press, pp. 265--283, 1998.

No context found.

C. Leacock and M. Chodorow. Combining local context and WordNet similarity for word sense identification. In C. Fellbaum, editor, WordNet: An electronic lexical database. MIT Press, 1998.

No context found.

C. Leacock and M. Chodorow, `Combining local context and wordnet similarity for word sense identification', in WordNet: An Electronic Lexical Database, ed., Christiane Fellbaum, 265--283, MIT Press, (1998).

No context found.

C. Leacock and M. Chodorow, `Combining local context and wordnet similarity for word sense identification', in WordNet: An Electronic Lexical Database, ed., Christiane Fellbaum, 265--283, MIT Press, (1998).

No context found.

Claudia Leacock e[Leacock&Chodorow 98] Claudia Leacock et Martin Chodorow. Combining Local Context and WordNet Similarity for Word Sense Identification. Christiane Fellbaum (ed.). WordNet: an electronic lexical database. Cambridge: MIT Press, pages 265-283.

No context found.

C. Leacock, M. Chodorow, Combining local context and WordNet similarity for word sense identification, in: C. Fellbaum (Ed.), WordNet: An electronic lexical database, MIT Press, 1998, pp. 265--283.

No context found.

Claudia Leacock and Martin Chodorow, Combining local context and WordNet similarity for word sense disambiguation, WordNet: an Electronic Lexical Database (Christiane Fellbaum, ed.), MIT Press, 1998, pp. 268--283.

No context found.

Leacock, C. and M. Chodorow, 1998, Combining Local Context and WordNet Similarity for Word Sense Identification, in: C. Fellbaum (editor) WordNet: An Electronic Lexical Database. pp. 265-283, The MIT Press: Cambridge, MA.

No context found.

Claudia Leacock and Martin Chodorow. 1998. Combining local context and WordNet similarity for word sense identification. In Fellbaum 1998, pp. 265--283.

No context found.

Language 73, 534-559. Leacock C and Chodorow M (1998). Combining Local Context and WordNet Similarity for Word Sense Identification. In: Fellbaum, C.

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