Sentiment Analysis Of Electricity Supplier Air Conditioning Products

On the network,the sentiment analysis of a large number of comment data is the latest category of Web mining,which has been widely concerned by various disciplines.With the rapid development of information technology,people are more inclined to use the Internet to buy their own love of things,there are a lot of Chinese business platform,online shopping behavior has brought opportunities Sheng’s major electricity supplier platform and the corresponding commodity manufacturer,but also a platform for the major manufacturers and brings challenges and so if you want to compete,each business growth in sales of goods,in addition to ensure genuine,but also need to understand the voices of consumers.Now more and more people choose shopping in the electronic business platform,many platforms are also corresponding with the comment system,users can view on the expression of commodity in the system,the user view has become an important basis for the development of electronic business platform and manufacturers.The text reviews data consumer sentiment analysis,can make manufacturers and business platform corresponding to the competitiveness is greatly improved,can develop new marketing strategies for the manufacturers and the corresponding business platform.It is an important work of this paper to use computer to help manufacturers and electronic business platform to mine and process the relevant text evaluation information.Three kinds of products of GREE brand air conditioner we choose Jingdong electricity supplier GREE KFR-35GW/(35592)KFR-72LW/(72596)FNhDa-A3,GREE FNAa-A3,GREE KFR-35GW/(35559)FNAd-A3(WIFI),mining of the three kinds of product reviews,to review data review information capture before July 11,2016 by crawler software-Octopus collector.Next to the data obtained for the basic data processing,the main text to weight,mechanical compression to the word,the Chinese review word segmentation,remove English and English characters,etc..The text reviews data after pretreatment of text data processing by the sentiment dictionary based on the emotional tendency of the text of the judgment and processing,thus the text data are divided into positive emotions and negative emotions of two parts;the comment data semantic network analysis,semantic network,data visualization;then on the three air conditioning products of the positive and negative comments of topic analysis based on LDA model,extracting three potential topics,looking for high-frequency feature comments in the data mining,the features of the products;at the end of the air conditioning product reviews data using tf/idf method to extract the attributes of goods,properties of three kinds of commodities to make comparison.Finally,through the analysis of the results obtained in the text of the text to explore the value of the content of the data,looking for three kinds of air conditioning,the most advantages and potential of the goods,to provide recommendations for Jingdong electricity supplier and GREE manufacturers.