Research On Calculation Method Of Semantic Praise Degree Based On Book Review Big Data

Books are the spiritual food of human beings.As the product of transmitting human wisdom and disseminating public knowledge,books are indispensable products of our human society.However,as the number of existing books is too large,the variety is various,and can't find the better books in similar books,coupled with the uneven quality,so people often have difficulty in choosing books and reading,and even can't make a reasonable judgment on the advantages and disadvantages of books.Therefore,if we can provide a method for book user to distinguish the book is "good or bad",it will be able to solve the puzzle of “difficult choices and distinguish”,so that the producers can produce good books,the sellers can sell good books,the readers can read good books.In order to evaluate the "book quality" more objectively,this paper proposes the concept of "semantic praise degree",and gives the calculation model of the degree of praise based on the large data of the book review.Firstly,the model is based on a large number of books evaluation information in the book sales website,by defining the rules of word segmentation,tagging the speech character,extracting the core keywords to realize the extraction of the book corpus,after filtering and deduplication,the valuable evaluation texts are retained.Then,the topic classification dictionary is built on this basis,and for the different topic categories constructed the book evaluation tendency dictionary,including the sentiment dictionary,the degree dictionary,and the negative dictionary.In order to achieve dynamic updates,data-driven methods are used to calculate the weight values of different dictionaries.In order to improve the utilization efficiency of corpus,a corpus focusing algorithm based on different topics was designed.Finally,according to the different types of topic comments,through the design of the single-sentence-level,user-level to book-level step by step,calculate the semantic praise degree,the algorithm based on rules and statistics is designed to calculate the semantic praise degree of books.This algorithm can evaluate the different topics of books.This paper designs and implements a semantic praise analysis system based on the big data of book reviews,and carries out experimental verification based on the book sales website.The experimental results show that the system can reflect the popularity of the book well and provide readers with more choices for books.v...