文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > Syntax versus Semantics Analysis of Enriched Vector Space Models
侵权投诉

Syntax versus Semantics Analysis of Enriched Vector Space Models

Abstract. This paper presents a robust method for the construction of collection-specific document models. These document models are variants of the well-known vector space model, which relies on a process of selecting, modifying, and weighting index terms

2

INDEXCONSTRUCTIONFORDOCUMENTMODELS

Thissectionorganizesthecurrentpracticeofindexconstructionforvectorspacemodels.Inparticular,wereviewtheconceptofadoc-umentmodelandproposeaclassi cationschemeforbothpopularandspecializedindexconstructionprinciples.

Adocumentdcanbeviewedunderdifferentaspects:layout,struc-turalorlogicalsetup,andsemantics.Acomputerrepresentationdofdmaycapturedifferentportionsofthesesaspects.Notethatdisdesignedpurposefully,withrespecttothestructureofaformalizedquery,q,andalsowithhavingaparticularretrievalmodelinmind.Aretrievalmodel,R,providesthelinguisticrationaleforthemodelfor-mationprocessbehindthemappingd→d.Thismappinginvolvesaninevitablesimpli cationofdthatshouldbe1.quanti able,

2.usefulwithrespecttotheinformationneed,and3.tailoredtoq,theformalizedquery.

TheretrievalmodelRgivesanswerstothesepoints,beittheo-reticallyorempirically,andprovidesaconcretemeans,ρ(q,d),forquantifyingtherelevancebetweenaformalizedqueryqandadocu-ment’scomputerrepresentationd.Notethatρ(q,d)isoftenspeci- edintheformofasimilaritymeasure .

Together,thecomputerrepresentationdalongwiththeunderlyingretrievalmodelRformthedocumentmodel;Figure2illustratestheconnections.

LetDbeadocumentcollectionandletTbethesetofalltermsthatoccurinD.Thevectorspacemodeldofadocumentdisavec-torof|T|weights,eachofwhichquantifyingthe“importance”ofsomeindexterminTwithrespecttod.3Thisquanti cationmustbeseenagainstthebackgroundthatoneisinterestedinasimilarityfunction thatmapsfromthevectorsd1andd2oftwodocumentsd1,d2intotheinterval[0;1]andthathasthefollowingproperty:If (d1,d2)iscloseto1thenthedocumentsd1andd2aresimilar;likewise,avalueclosetozeroindicatesahighdissimilarity.Notethatdocumentmodelsandsimilarityfunctionsdetermineeachother:Thevectorspacemodelanditsvariantsareamenabletothecosinesimi-larity(=normalizeddotproduct)in rstplace,butcanalsobeusedinconnectionwithEuclideandistance,overlapmeasures,orotherdistanceconcepts.

Underthevectorspaceparadigmthedocumentmodelconstruc-tionprocessisdeterminedintwodimensions:indexconstructionandweightcomputation.Inthefollowingwewillconcentrateonthefor-merdimensionsincethispapercontributesrighthere.Wehaveclas-

上一页第2页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email