电子学报
電子學報
전자학보
ACTA ELECTRONICA SINICA
2014年
4期
740-746
,共7页
在线产品评论%满意度评价%灰色理论%观点挖掘%相似性情感词匹配词对%对应性热点词匹配词对
在線產品評論%滿意度評價%灰色理論%觀點挖掘%相似性情感詞匹配詞對%對應性熱點詞匹配詞對
재선산품평론%만의도평개%회색이론%관점알굴%상사성정감사필배사대%대응성열점사필배사대
online product review%satisfaction degree evaluation%grey theory%opinion mining%matching word pair of similar sentiment words%corresponding top word pair
文中提出一种基于灰色理论的用户满意度评价方法。基本思想:根据对各种类型的产品和对象(如旅社、酒店等)的评价具有相对性、不确定性和模糊性(这三种特性统称为灰色性)的特征,将用户满意度分为很满意,满意,一般,不满意和很不满意等五种类型,克服了已有方法将情感分为满意和不满意的粗粒度定性评价缺陷。构建了用户满意度灰色评估模型,为了扩大对各类型产品评价的适应性,引用了相似性情感词匹配词对和对象对应产品的对应性热点词匹配词对来构建用户对产品和对象的满意度评测指标。以实际产品为实验数据对象,仿真结果表明:提出的方法不仅能获得产品或对象的总体满意度结论值 VB ,还能得到各项评测指标(影响产品或对象的因素)的量化评估结论值 VB i ,有效解决了定性评价方法的不足,为消费者提供了更详细的消费指导,为商业机构提供了有针对性的生产销售措施。
文中提齣一種基于灰色理論的用戶滿意度評價方法。基本思想:根據對各種類型的產品和對象(如旅社、酒店等)的評價具有相對性、不確定性和模糊性(這三種特性統稱為灰色性)的特徵,將用戶滿意度分為很滿意,滿意,一般,不滿意和很不滿意等五種類型,剋服瞭已有方法將情感分為滿意和不滿意的粗粒度定性評價缺陷。構建瞭用戶滿意度灰色評估模型,為瞭擴大對各類型產品評價的適應性,引用瞭相似性情感詞匹配詞對和對象對應產品的對應性熱點詞匹配詞對來構建用戶對產品和對象的滿意度評測指標。以實際產品為實驗數據對象,倣真結果錶明:提齣的方法不僅能穫得產品或對象的總體滿意度結論值 VB ,還能得到各項評測指標(影響產品或對象的因素)的量化評估結論值 VB i ,有效解決瞭定性評價方法的不足,為消費者提供瞭更詳細的消費指導,為商業機構提供瞭有針對性的生產銷售措施。
문중제출일충기우회색이론적용호만의도평개방법。기본사상:근거대각충류형적산품화대상(여려사、주점등)적평개구유상대성、불학정성화모호성(저삼충특성통칭위회색성)적특정,장용호만의도분위흔만의,만의,일반,불만의화흔불만의등오충류형,극복료이유방법장정감분위만의화불만의적조립도정성평개결함。구건료용호만의도회색평고모형,위료확대대각류형산품평개적괄응성,인용료상사성정감사필배사대화대상대응산품적대응성열점사필배사대래구건용호대산품화대상적만의도평측지표。이실제산품위실험수거대상,방진결과표명:제출적방법불부능획득산품혹대상적총체만의도결론치 VB ,환능득도각항평측지표(영향산품혹대상적인소)적양화평고결론치 VB i ,유효해결료정성평개방법적불족,위소비자제공료경상세적소비지도,위상업궤구제공료유침대성적생산소수조시。
This paper proposed a kind of method to evaluate customer satisfaction degree based on grey theory .Core idea is that the user satisfaction degree is divided into five categories as very satisfied ,satisfied ,ok ,unsatisfied and very unsatisfied accord-ing to the comments ,which are characterized by the relativity ,uncertainty and ambiguity(collectively called Grey trait) ,from variety of products and objects(e .g .hotels ,restaurants) .The proposed method addressed the limit of the common method which only parti -tions qualitatively the customer sentiment into two coarse-grained categories :satisfied and unsatisfied .A grey model is constructed for evaluating customer satisfaction degree .In order to augment the applicability to evaluate the various types of products ,the match-ing word pair of similar sentiment words and the corresponding top word pair of the object mapping the product are introduced to build the evaluation index by the customer satisfaction degree for the product and object .Experiments are conducted on the real life dataset ,and the results show that it not only obtains the overall satisfaction degree ,but also the quantitative evaluation value of the computed affecting factors of the product and the object .So the proposed method would effectively overcome the disadvantage of qualitative evaluation method ,provide detailed guidance to consumers ,and offer targeted marketing techniques for commercial orga-nizations .