?url_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rft.title=Coherent+Keyphrase+Extraction+via+Web+Mining&rft.creator=Turney%2C+Peter&rft.subject=Statistical+Models&rft.subject=Language&rft.subject=Machine+Learning&rft.description=Keyphrases+are+useful+for+a+variety+of+purposes%2C%0Aincluding+summarizing%2C+indexing%2C+labeling%2C%0Acategorizing%2C+clustering%2C+highlighting%2C+browsing%2C+and%0Asearching.+The+task+of+automatic+keyphrase+extraction%0Ais+to+select+keyphrases+from+within+the+text+of+a+given%0Adocument.+Automatic+keyphrase+extraction+makes+it%0Afeasible+to+generate+keyphrases+for+the+huge+number+of%0Adocuments+that+do+not+have+manually+assigned%0Akeyphrases.+A+limitation+of+previous+keyphrase%0Aextraction+algorithms+is+that+the+selected+keyphrases+are%0Aoccasionally+incoherent.+That+is%2C+the+majority+of+the%0Aoutput+keyphrases+may+fit+together+well%2C+but+there+may%0Abe+a+minority+that+appear+to+be+outliers%2C+with+no+clear%0Asemantic+relation+to+the+majority+or+to+each+other.+This%0Apaper+presents+enhancements+to+the+Kea+keyphrase%0Aextraction+algorithm+that+are+designed+to+increase+the%0Acoherence+of+the+extracted+keyphrases.+The+approach+is%0Ato+use+the+degree+of+statistical+association+among%0Acandidate+keyphrases+as+evidence+that+they+may+be%0Asemantically+related.+The+statistical+association+is%0Ameasured+using+web+mining.+Experiments+demonstrate%0Athat+the+enhancements+improve+the+quality+of+the%0Aextracted+keyphrases.+Furthermore%2C+the+enhancements%0Aare+not+domain-specific%3A+the+algorithm+generalizes+well%0Awhen+it+is+trained+on+one+domain+(computer+science%0Adocuments)+and+tested+on+another+(physics+documents).&rft.date=2003&rft.type=Conference+Paper&rft.type=PeerReviewed&rft.format=application%2Fpdf&rft.identifier=http%3A%2F%2Fcogprints.org%2F3122%2F1%2FNRC-46496.pdf&rft.identifier=++Turney%2C+Peter++(2003)+Coherent+Keyphrase+Extraction+via+Web+Mining.++%5BConference+Paper%5D+++++&rft.relation=http%3A%2F%2Fcogprints.org%2F3122%2F