Vector Space Models of Word Meaning and Phrase Meaning: A Survey
Identifieur interne : 000205 ( Main/Exploration ); précédent : 000204; suivant : 000206Vector Space Models of Word Meaning and Phrase Meaning: A Survey
Auteurs : Katrin Erk [États-Unis]Source :
- Language and Linguistics Compass [ 1749-818X ] ; 2012-10.
English descriptors
- Teeft :
- Annual meeting, Author language, Baroni, Bayesian approaches, Blackwell, Blackwell publishers, Blackwell publishing, Cambridge university press, Cognitive, Cognitive science, Cognitive science society, Computational, Computational linguistics, Computational semantics, Conceptual spaces, Context items, Context word, Context words, Contextual similarity, Corpus data, Data structures, Dictionary senses, Different senses, Dimensionality reduction, Distributional, Distributional approaches, Distributional context, Distributional hypothesis, Distributional model, Distributional models, Distributional representations, Distributional semantics, Distributional similarity, Dumais, Empirical methods, Feature norms, General framework, Geometrical models, Human subjects, Indirect supervision, Individual word occurrences, Inference rules, Information retrieval, International conference, Intransitive verb, July, Katrin, Landauer, Language model, Lapata, Linear algebra, Linguistics, Linguistics compass, Many word meaning phenomena, Matrix, Mental concepts, Modeling, Natural language processing, Natural language semantics, Noun, Other approaches, Pado, Particular sentence context, Phrase meaning, Phrase models, Phrase representations, Phrase similarity, Phrase vectors, Polysemy, Saarland university, Sadrzadeh, Same structure, Sample sentences, Selectional preferences, Semantic, Semantic composition, Semantic relations, Semantic representation, Semantic similarity, Semantic space, Semantic types, Semantics, Sens, Sentence context, Sentence semantics, Sentence structure, Similar contexts, Similar words, Similarity, Similarity measure, Similarity predictions, Simplest case, Special interest group, Strawberry, Such models, Symbol interdependency, Syntactic, Syntactic relation, Syntactic structure, Target word, Target words, Tensor, Tensor product, Topic model, Tough problem, Typical arguments, Vector, Vector addition, Vector combination, Vector space, Vector space model, Vector space models, Vector space representation, Vector space representations, Vector spaces, Word meaning, Word meaning representations, Word occurrence vectors, Word sense disambiguation models, Word senses, Word similarity, Word vector, Word vectors.
Abstract
Distributional models represent a word through the contexts in which it has been observed. They can be used to predict similarity in meaning, based on the distributional hypothesis, which states that two words that occur in similar contexts tend to have similar meanings. Distributional approaches are often implemented in vector space models. They represent a word as a point in high‐dimensional space, where each dimension stands for a context item, and a word's coordinates represent its context counts. Occurrence in similar contexts then means proximity in space. In this survey we look at the use of vector space models to describe the meaning of words and phrases: the phenomena that vector space models address, and the techniques that they use to do so. Many word meaning phenomena can be described in terms of semantic similarity: synonymy, priming, categorization, and the typicality of a predicate's arguments. But vector space models can do more than just predict semantic similarity. They are a very flexible tool, because they can make use of all of linear algebra, with all its data structures and operations. The dimensions of a vector space can stand for many things: context words, or non‐linguistic context like images, or properties of a concept. And vector space models can use matrices or higher‐order arrays instead of vectors for representing more complex relationships. Polysemy is a tough problem for distributional approaches, as a representation that is learned from all of a word's contexts will conflate the different senses of the word. It can be addressed, using either clustering or vector combination techniques. Finally, we look at vector space models for phrases, which are usually constructed by combining word vectors. Vector space models for phrases can predict phrase similarity, and some argue that they can form the basis for a general‐purpose representation framework for natural language semantics.
Url:
DOI: 10.1002/lnco.362
Affiliations:
Links toward previous steps (curation, corpus...)
- to stream Istex, to step Corpus: 000995
- to stream Istex, to step Curation: 000943
- to stream Istex, to step Checkpoint: 000075
- to stream Main, to step Merge: 000205
- to stream Main, to step Curation: 000205
Le document en format XML
<record><TEI wicri:istexFullTextTei="biblStruct"><teiHeader><fileDesc><titleStmt><title xml:lang="en">Vector Space Models of Word Meaning and Phrase Meaning: A Survey</title>
<author><name sortKey="Erk, Katrin" sort="Erk, Katrin" uniqKey="Erk K" first="Katrin" last="Erk">Katrin Erk</name>
</author>
</titleStmt>
<publicationStmt><idno type="wicri:source">ISTEX</idno>
<idno type="RBID">ISTEX:5DBCE29AF41362B802BE96E5BDB1392ADFF7CECE</idno>
<date when="2012" year="2012">2012</date>
<idno type="doi">10.1002/lnco.362</idno>
<idno type="url">https://api.istex.fr/document/5DBCE29AF41362B802BE96E5BDB1392ADFF7CECE/fulltext/pdf</idno>
<idno type="wicri:Area/Istex/Corpus">000995</idno>
<idno type="wicri:explorRef" wicri:stream="Istex" wicri:step="Corpus" wicri:corpus="ISTEX">000995</idno>
<idno type="wicri:Area/Istex/Curation">000943</idno>
<idno type="wicri:Area/Istex/Checkpoint">000075</idno>
<idno type="wicri:explorRef" wicri:stream="Istex" wicri:step="Checkpoint">000075</idno>
<idno type="wicri:doubleKey">1749-818X:2012:Erk K:vector:space:models</idno>
<idno type="wicri:Area/Main/Merge">000205</idno>
<idno type="wicri:Area/Main/Curation">000205</idno>
<idno type="wicri:Area/Main/Exploration">000205</idno>
</publicationStmt>
<sourceDesc><biblStruct><analytic><title level="a" type="main" xml:lang="en">Vector Space Models of Word Meaning and Phrase Meaning: A Survey</title>
<author><name sortKey="Erk, Katrin" sort="Erk, Katrin" uniqKey="Erk K" first="Katrin" last="Erk">Katrin Erk</name>
<affiliation wicri:level="4"><country>États-Unis</country>
<placeName><settlement type="city">Austin (Texas)</settlement>
<region type="state">Texas</region>
</placeName>
<orgName type="university">Université du Texas à Austin</orgName>
</affiliation>
</author>
</analytic>
<monogr></monogr>
<series><title level="j">Language and Linguistics Compass</title>
<idno type="ISSN">1749-818X</idno>
<idno type="eISSN">1749-818X</idno>
<imprint><publisher>Blackwell Publishing Ltd</publisher>
<pubPlace>Oxford, UK</pubPlace>
<date type="published" when="2012-10">2012-10</date>
<biblScope unit="volume">6</biblScope>
<biblScope unit="issue">10</biblScope>
<biblScope unit="page" from="635">635</biblScope>
<biblScope unit="page" to="653">653</biblScope>
</imprint>
<idno type="ISSN">1749-818X</idno>
</series>
<idno type="istex">5DBCE29AF41362B802BE96E5BDB1392ADFF7CECE</idno>
<idno type="DOI">10.1002/lnco.362</idno>
<idno type="ArticleID">LNC3362</idno>
</biblStruct>
</sourceDesc>
<seriesStmt><idno type="ISSN">1749-818X</idno>
</seriesStmt>
</fileDesc>
<profileDesc><textClass><keywords scheme="Teeft" xml:lang="en"><term>Annual meeting</term>
<term>Author language</term>
<term>Baroni</term>
<term>Bayesian approaches</term>
<term>Blackwell</term>
<term>Blackwell publishers</term>
<term>Blackwell publishing</term>
<term>Cambridge university press</term>
<term>Cognitive</term>
<term>Cognitive science</term>
<term>Cognitive science society</term>
<term>Computational</term>
<term>Computational linguistics</term>
<term>Computational semantics</term>
<term>Conceptual spaces</term>
<term>Context items</term>
<term>Context word</term>
<term>Context words</term>
<term>Contextual similarity</term>
<term>Corpus data</term>
<term>Data structures</term>
<term>Dictionary senses</term>
<term>Different senses</term>
<term>Dimensionality reduction</term>
<term>Distributional</term>
<term>Distributional approaches</term>
<term>Distributional context</term>
<term>Distributional hypothesis</term>
<term>Distributional model</term>
<term>Distributional models</term>
<term>Distributional representations</term>
<term>Distributional semantics</term>
<term>Distributional similarity</term>
<term>Dumais</term>
<term>Empirical methods</term>
<term>Feature norms</term>
<term>General framework</term>
<term>Geometrical models</term>
<term>Human subjects</term>
<term>Indirect supervision</term>
<term>Individual word occurrences</term>
<term>Inference rules</term>
<term>Information retrieval</term>
<term>International conference</term>
<term>Intransitive verb</term>
<term>July</term>
<term>Katrin</term>
<term>Landauer</term>
<term>Language model</term>
<term>Lapata</term>
<term>Linear algebra</term>
<term>Linguistics</term>
<term>Linguistics compass</term>
<term>Many word meaning phenomena</term>
<term>Matrix</term>
<term>Mental concepts</term>
<term>Modeling</term>
<term>Natural language processing</term>
<term>Natural language semantics</term>
<term>Noun</term>
<term>Other approaches</term>
<term>Pado</term>
<term>Particular sentence context</term>
<term>Phrase meaning</term>
<term>Phrase models</term>
<term>Phrase representations</term>
<term>Phrase similarity</term>
<term>Phrase vectors</term>
<term>Polysemy</term>
<term>Saarland university</term>
<term>Sadrzadeh</term>
<term>Same structure</term>
<term>Sample sentences</term>
<term>Selectional preferences</term>
<term>Semantic</term>
<term>Semantic composition</term>
<term>Semantic relations</term>
<term>Semantic representation</term>
<term>Semantic similarity</term>
<term>Semantic space</term>
<term>Semantic types</term>
<term>Semantics</term>
<term>Sens</term>
<term>Sentence context</term>
<term>Sentence semantics</term>
<term>Sentence structure</term>
<term>Similar contexts</term>
<term>Similar words</term>
<term>Similarity</term>
<term>Similarity measure</term>
<term>Similarity predictions</term>
<term>Simplest case</term>
<term>Special interest group</term>
<term>Strawberry</term>
<term>Such models</term>
<term>Symbol interdependency</term>
<term>Syntactic</term>
<term>Syntactic relation</term>
<term>Syntactic structure</term>
<term>Target word</term>
<term>Target words</term>
<term>Tensor</term>
<term>Tensor product</term>
<term>Topic model</term>
<term>Tough problem</term>
<term>Typical arguments</term>
<term>Vector</term>
<term>Vector addition</term>
<term>Vector combination</term>
<term>Vector space</term>
<term>Vector space model</term>
<term>Vector space models</term>
<term>Vector space representation</term>
<term>Vector space representations</term>
<term>Vector spaces</term>
<term>Word meaning</term>
<term>Word meaning representations</term>
<term>Word occurrence vectors</term>
<term>Word sense disambiguation models</term>
<term>Word senses</term>
<term>Word similarity</term>
<term>Word vector</term>
<term>Word vectors</term>
</keywords>
</textClass>
<langUsage><language ident="en">en</language>
</langUsage>
</profileDesc>
</teiHeader>
<front><div type="abstract" xml:lang="en">Distributional models represent a word through the contexts in which it has been observed. They can be used to predict similarity in meaning, based on the distributional hypothesis, which states that two words that occur in similar contexts tend to have similar meanings. Distributional approaches are often implemented in vector space models. They represent a word as a point in high‐dimensional space, where each dimension stands for a context item, and a word's coordinates represent its context counts. Occurrence in similar contexts then means proximity in space. In this survey we look at the use of vector space models to describe the meaning of words and phrases: the phenomena that vector space models address, and the techniques that they use to do so. Many word meaning phenomena can be described in terms of semantic similarity: synonymy, priming, categorization, and the typicality of a predicate's arguments. But vector space models can do more than just predict semantic similarity. They are a very flexible tool, because they can make use of all of linear algebra, with all its data structures and operations. The dimensions of a vector space can stand for many things: context words, or non‐linguistic context like images, or properties of a concept. And vector space models can use matrices or higher‐order arrays instead of vectors for representing more complex relationships. Polysemy is a tough problem for distributional approaches, as a representation that is learned from all of a word's contexts will conflate the different senses of the word. It can be addressed, using either clustering or vector combination techniques. Finally, we look at vector space models for phrases, which are usually constructed by combining word vectors. Vector space models for phrases can predict phrase similarity, and some argue that they can form the basis for a general‐purpose representation framework for natural language semantics.</div>
</front>
</TEI>
<affiliations><list><country><li>États-Unis</li>
</country>
<region><li>Texas</li>
</region>
<settlement><li>Austin (Texas)</li>
</settlement>
<orgName><li>Université du Texas à Austin</li>
</orgName>
</list>
<tree><country name="États-Unis"><region name="Texas"><name sortKey="Erk, Katrin" sort="Erk, Katrin" uniqKey="Erk K" first="Katrin" last="Erk">Katrin Erk</name>
</region>
</country>
</tree>
</affiliations>
</record>
Pour manipuler ce document sous Unix (Dilib)
EXPLOR_STEP=$WICRI_ROOT/Wicri/Sarre/explor/MusicSarreV3/Data/Main/Exploration
HfdSelect -h $EXPLOR_STEP/biblio.hfd -nk 000205 | SxmlIndent | more
Ou
HfdSelect -h $EXPLOR_AREA/Data/Main/Exploration/biblio.hfd -nk 000205 | SxmlIndent | more
Pour mettre un lien sur cette page dans le réseau Wicri
{{Explor lien |wiki= Wicri/Sarre |area= MusicSarreV3 |flux= Main |étape= Exploration |type= RBID |clé= ISTEX:5DBCE29AF41362B802BE96E5BDB1392ADFF7CECE |texte= Vector Space Models of Word Meaning and Phrase Meaning: A Survey }}
This area was generated with Dilib version V0.6.33. |