A Study On The Relationship Between Online Search Data,Online Word-of-Mouth And Chinese Automobile Sales

With the widespread use of search engines,the sources of consumer information have gradually shifted from traditional media such as newspapers,magazines,radio and television,and physical stores to Internet search channels.The search engine records the user's search behavior.The network search data reflects the user's attention to the event.The network search data,which provides low cost and timeliness,has provided many help for all walks of life and has been widely used in the social economy.In the prediction of behavior.Web search behavior is a measure of consumer's willingness to buy and can be used to predict purchase behavior.Although information search has always been regarded as an important part of the consumer decision-making process,there is limited empirical research on the relationship between online search and offline transactions.Based on this,this article takes the actual data left by consumers in the process of information collection and purchase decision-making as the research basis,and takes China's auto market as the background to explore the relationship between online search and offline transactions.First of all,to determine the keywords of web search data,text mining techniques are mainly used.Specifically: 1 Carry out the Jieba word segmentation for the scraped car forum texts;2 Use Word2 vec model to convert the word segmentation results into a vector space model;3 Combine TF The IDF algorithm and the cosine similarity algorithm determine the keywords.Then,based on 108 months of long panel data,a fixed effect model of web search and car sales was established.In addition,in order to explore the effect of online word-of-mouth on auto sales,this paper introduces two variables,namely comment number and comment score,based on the original model,to test the relationship between online search,online word-of-mouth and Chinese auto sales.The retention sample analysis shows that when online word-of-mouth information is unavailable,online search data has played an important role in predicting car sales.Finally,using the panel-vector autoregressive model(PVAR),a rolling window approach was used to use the web search data to forecast vehicle sales for the 12 months of 2015.The empirical results of this paper show that: 1.There is a long-term equilibrium relationship between online search and car sales,and the regression model can explain76% of the variance;2.The number of car online reviews and the average emotional tendency have a significant positive effect on consumers' purchase behavior.Impact;3.Online word-of-mouth data can't play a role in the sales forecasting model based on web search data;4.Prediction models based on web search index show good prediction accuracy and can be used one month or two months in advance.Search data on the Internet,forecasting China's car sales.The research of this paper enriches the theoretical researches on web search,online word-of-mouth,automobile marketing and other related aspects,which is helpful for automobile companies to formulate relevant marketing strategies.In addition,the keyword acquisition process of this article also provides reference for the research of web search data in other fields.