Feedback Properties Of Goods Automatic Annotation System Design And Implementation

With the development of B2B networks, online shopping has become a new kind ofconsumption patterns, which allows efficient B2B, at different times, in differentlocations, possible.At present, the network has accumulated a large number of Web commodityresources, but due to the simple annotation structure of comment properties of Webcommodity resources, the lack of semantics, and especially the serious deficiencies inautomatic annotation of comment properties of Web commodity resources, there are notimely advice of reference for users, which restricts the development of B2B networks.Therefore, how to annotate the Web commodity resources with comment attributes, toprovide feedback to the users of advice about the resources and to provide users withauto-tagging resource views would make related resources information easily andquickly presented to the users has become urgent problems of network B2B.This thesis introduces comment vacabulary Ontology techniques and commentvacabulary similarity computing model to comment properties of Web commodityautomatic annotation system. It uses the formal theory of Ontology in quotient space tobuilt comment vocabulary Ontology and realizes random jump and reasoning ofcomment vocabulary on semantic level. By using comment vocabulary similaritycalculation model to calculate the comment meta-vocabulary and the similarity ofcomment vocabulary Ontology to determine the satisfaction degree of the commodityresources, it implements the semantic automatic annotation systems of commodityresources comments, which provides users with the general opinions of commodityresources to facilitate the users’ query. The main content of this thesis are:(1) Combined with the software development process in software engineering, itproposes the formal theory of Ontology in quotient space, which make random jumpand reasoning of comment vocabulary Ontology in construction levels possible. Bycombining with the basic modeling methods of Ontology and development tools Hozo,it could create knowledge model of comment vocabulary Ontology and provide classification system of comment vocabularies.(2) Through deep understanding of the and the hierarchy structrue of Ontology, itbuilds the semantic similarity calculation model under concept lattice theory.Meanwhile, it analyzes the comment meta-vocabulary and comment vocabularyOntology to compute the similarity of comment meta-vocabulary and commentvocabulary Ontology, in order to provide automatic semantic annotation of commentproperty of commodity resources and to form the comment description document of thecommodity resources..(3) Based on the constructed comment vocabulary Ontology knowledge and thesemantic similarity computing model, it uses programming language Java to implementa prototype system, which could extract the comment meta-vocabularies of Webcommodity resources and calculate the similarity. Through calculation of the similarityof comment meta-vocabulary, this system can render the user automatic annotation ofneeded comment of the commodity resources and provide online shopping advice forusers reference.The core issue of this thesis is the construction of opinion vocabulary ontology andhow to calculate the similarity computing model of the comment meta-vocabulary andcomment vocabulary Ontology. The comment vocabulary Ontology can not onlyprovide better comment classification structure of the Web commodity resources, andcan also be applied to other areas. Therefore, it has a very good research value.