Emotional Inclination Analysis Of E-commerce Online Commentary

In recent years,the development of e-commerce has greatly promoted people's enthusiasm for shopping through the Internet.From the process of searching and discovering,browsing product details,adding to shopping carts,paying,receiving goods,and generating shopping reviews,they can complete without leaving the house..Among them,product reviews have become one of the important reference contents for consumers when purchasing products,because the generated comments can make consumers more comprehensively understand whether the products are worth buying from the perspective of merchants and advertisements.Similarly,manufacturers and ecommerce platforms need to pay close attention to the voice of consumers,and adjust product design or marketing strategies in a timely manner to better serve consumers while maximizing profits.Based on this,this paper uses the data acquisition,cleaning,extraction of product feature words and sentiment words,triad text vectorization and SVM sentiment classification to form a PROST model for the 4000 comments of the Jingdong Ecommerce platform mobile phone.Extracting the characteristics of consumers' most concerned about the goods and extracting their corresponding sentimental tendencies,and comparing the new model with the traditional methods,the conclusion that the new method is superior to the traditional method is obtained,and the sentiment orientation analysis of e-commerce online comment texts is analyzed.Provides a simple and feasible method of operation.The main writing ideas of this paper are divided into the following parts:The first part is mainly the introduction and basic theoretical knowledge.It introduces the research background and research significance of this paper,summarizes the progress of the analysis of the sentiment orientation of domestic and foreign texts,and combs the research content,research methods and technical routes used and innovation.point.The second part introduces three theoretical supports,namely,ecommerce related theory,sentiment analysis related theory,text analysis related theory,and e-commerce online commentary text sentiment orientation analysis.The third part introduces the text acquisition and preprocessing,and systematically expounds the text segmentation method,part-of-speech tagging and syntactic dependency.The fourth part introduces the common feature vectorization method and sentiment analysis model,emphasizes the necessity and importance of text vectorization,and selects the CBOW model that can obtain the relationship between words to vectorize the text.The fifth part is the experimental process and the comparative analysis of the results.The text data of the experimental process is obtained and cleaned,and the word segmentation and the dependent syntax are used to extract the triplet and carry out the continuous word bag model vectorization and classifier classification.The experimental results obtained are compared with the four traditional methods,and the conclusion that the performance of the new model is better than the traditional model is obtained.The sixth part is the summary and outlook.It discusses and analyzes the pros and cons of the PROST model provided in this paper,and proposes the future research direction.