Domain Ontology Construction And Its Application In Comment Target Extraction Of Prodcut Reviews

In today's society,online shopping has become an important part of people's daily life.The product review function provided by the e-commerce platform allows consumers to share the shopping experience and comment on the purchased products or services.These comments have become an important resource for potential consumers to make purchasing decisions,and provide an important basis for companies to develop marketing strategies and optimize products and services.However,the era of big data brings information overload,coupled with the unstructured nature of product reviews,which makes it difficult for companies and consumers to quickly capture valid information in reviews.In this context,the opinion mining of product reviews came into being,meeting the needs of information acquisition in online trading activities.As the primary task of opinion mining,the accurate extraction of the comment target will lay the foundation for the validity and scientificity of the follow-up research.At present,in the field of product comment target extraction,most scholars pay attention to how to improve the existing explicit comment target extraction algorithm,but does not pay much attention to the recognition of inter-concept relationship in the explicit comment target extraction result and the implicit comment target extraction,which leads to that the product comment target extraction still has the problems such as the low degree accuracy of comment target recognition,and redundant extraction results,lack of relevance,etc.In order to improve the effect of product comment target extraction,this paper attempts to explore the support and role provided by the introduction of domain ontology.Based on the related ontology theory,the knowledge of product review domain is modeled for comment target extraction.The classes,attributes and instances in ontology are clearly defined,and evaluate the constructed product review ontology.We use the existing association rule algorithm as the underlying algorithm,combined with the semantic analysis and rule reasoning method to develop a complete set of comment target extraction implementation steps and methods,using the semantic knowledge of ontology realizes the automatic acquisition of the relationship between concepts in the extraction result of the explicit comment target.Based on the identification relationship between the comment target and the identification word,the implicit comment target in the product review is automatically extracted by the ontology reasoning mechanism.The effect of the comment target extraction mechanism proposed in this paper is evaluated through experiments.The conclusion of this paper is that:We construct the ontology model for the product domain knowledge and introduce the subjective product review information into it,which can effectively improve the reliability of comment target extraction.In explicit comment target extraction,combined with semantic similarity theory,the extraction result is synonymous with the ontology concept,which can realize the standardization of the extraction result,reduce the redundancy,and automatically acquire the relationship between the concepts in the extraction result by relying on the ontology knowledge,which can enhance the correlation between concepts.In the implicit comment target extraction,based on the identification relationship between the comment target and the identification word,We use ontology reasoning to identify the implicit comment target in the product review,which can satisfy the intelligent requirement of the automatic extraction of the implicit comment target.The evaluation of the effect of the comment target extraction through experiments can verify the accuracy of the established comment target extraction mechanism.The contribution of this paper shows that:Combined with ontology theory,We construct the structural model for product review domain knowledge,and a detailed description of the product review ontology construction process through examples.Based on association rules and semantic analysis techniques,specify the steps and methods of obtaining the inter-concept relationship in the extraction result of the explicit comment target.The relevant rules for implicit comment target extraction are defined in the updated product review ontology,which is used as the foundation of the follow-up implicit comment target inference implementation.Combined with the experimental results to evaluate the effect of the comment target extraction mechanism,to provide relevant enterprises with solutions for higher-efficiency product review information services,and provide a new ideas for the academic community to explore the intelligent technology of product review objects.