基于多值詞匯語義相似度的SME模型改進(jìn).pdf

1、河南大學(xué)碩士學(xué)位論文基于多值詞匯語義相似度的SME模型改進(jìn)姓名：周倩申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：教育技術(shù)學(xué)指導(dǎo)教師：沈夏炯201105IIIABSTRACTStructureMappingEngineSMEisananalogyinferencecalculationmodelwhichisproposedbasedonstructuralMappingTheySMT.Itismainlyknownfromtheknowledgethatb

2、etweentheoldknowledgenewknowledgetofindamostsimilaroneusethesimilaritybetweentheoldnewknowledgestructureestablishingtheconnectionbetweentheoldnewknowledge.Thelearnerscanlearnnewknowledgebyusetheknownknowledgetheconnectio

3、nbetweenthenewknowledge.TherearetwoproblemsinSMEanalogyinferencecalculationmodelthefirstisthatthenewknowledgestructuretheknowledgestructuretheneedofhumangivenitmakesgenerateknowledgestructureprocesstroubleinaccuratethese

4、condistocalculatethesemanticsimilarityofwdstwovalues01itmakestheanalogyofthelastgenerationcannotdeterminetheusecasescannotevenproduceananalogusecases.Inviewoftheabovetwoproblemsthispapermainlybasedonresearchtherelational

5、datasourcesitwillbasedon“Knowledgewk“inthefmofsemanticsimilaritycalculationthebackgroundaddedtotheanalogywiththestructuremodelwithafmalcontextthatthestructureofknowledgegenerateimprovedmodel.Themainresearchisdoneinthispa

6、percanbeclassedasfollows:FirstlyThesemanticsimilaritycalculationmethodwhichbasedon“Knowledgewkt“isappliedinanalogicalreasoningcalculationmodel.Thetraditionalmodelofanalogicalreasoningtocalculatethevalueoflexicalsemantics