Research On Online Review Mining Of Cruise Tourism Based On Domain Ontology

The in-depth development of the Internet has changed the habits of consumers.Online consumption and user reviews have become the common behavior of consumers.Online user reviews are increasing rapidly,and users express their opinions and feelings through comments.This provides fertile soil for Natural Language Processing's development.With the continuous development of natural language processing,opinion mining technology is becoming more and more mature.In order to mine the user's opinion emotion and clarify the user's consumption focus,it is necessary to analyze and mine the user's online comment.Users' online comments cover a wide range of fields,among which tourism is a special field,different from other tangible products,and tourism belongs to service products.Especially for cruise tourism,the consumer group is small,facing the middle and high-end groups,and its user evaluation is difficult to obtain through the original methods such as survey method,so its user online evaluation is very precious.Therefore,this paper combs the methods of online comment mining,constructs an online comment mining model,and applies it to the field of cruise tourism to clarify the user evaluation emotion in the field of cruise tourism,which is of great significance both in theory and practice.Based on domain ontology,this paper constructs an online review mining model of cruise tourism.Firstly,domain ontology is constructed to clarify the scope of research domain and improve the accuracy of the final emotional analysis results.The domain ontology is constructed by seven steps,the domain category and reuse possibility are determined,the domain concept words and terms,the definition class and hierarchy,the definition attributes and their facets are defined,and the examples are determined.Finally,the domain ontology hierarchy chart,the class and examples are obtained.Secondly,emotion dictionary is constructed,which is divided into three parts: basic emotion dictionary,domain emotion dictionary and supplementary emotion dictionary.The basic affective dictionary uses the existing and more commonly used HowNet affective words,and the domain affective dictionary uses SO-PMI algorithm to expand the affective words to get targeted domain affective words,while the supplementary affective dictionary mainly considers the impact of negative affective words,degree adverbs and emoticons on users' evaluation of emotion.Finally,the emotion calculation of user comments is carried out.The algorithm is used to calculate the overall emotion of user comments.The feature words of user comments are extracted by combining CountVectorizer and LDA.The final feature words are determined by combining the classes and examples of domain ontology to define the main object of emotion calculation.The last,a high accuracy online review mining model of cruise tourism based on domain ontology is obtained.The model is applied to the field of cruise tourism experiments,and it is clear that the model has a high accuracy.This research successfully constructs online opinion mining model,calculates and analyzes user comments in cruise tourism field,the following research conclusions are finally obtained: on the theoretical level,the online comment mining model based on domain ontology can accurately calculate the online comment emotion.In the field of cruise tourism,the accuracy is 90.3%,which proves the feasibility of this method.At the practical level,the model is applied to the field of cruise tourism.The experimental results show that:(1)most of the users in the cruise tourism market have a positive evaluation on cruise tourism,and the best evaluation on cruise accommodation,and the worst evaluation on cruise catering.(2)The "Matthew effect" and " Pareto‘s principle" of cruise tourism market are significant.(3)The cruise brand positioning is clear and the user coverage is broad.But the sales channel is relatively single.