Research On Product Sorting Of Online Short Rental Markets Based On Comment Mining

With the popularity of the online short-term rental market and the excessive expansion of online commentary information,it is difficult for consumers to judge whether the product meets their expectations through online commentary information before purchasing decisions.Especially in the field of online short-term rental,consumers buy the right to use the goods for a certain period of time,and experience the services of the merchants during this time,but it is difficult for consumers to get experience from the information provided by the merchants.So,how do consumers make choices? The problem raised in this paper is that consumers can satisfy basic needs through various basic information provided by sellers,and on the premise that the online reviews of these products are similar,how can consumers follow the information of online reviews? The products are further compared to select better products in an objective sense.This thesis combines the five-stage model of consumer purchasing decisionmaking,elaboration likelihood model and the information adoption model,combined with the inherent characteristics of the online short-term rental market,simplifies the discussion of factors with consensus conclusions,namely the individual characteristics of the reviewers on the edge path,and focuses on the central path.On the analysis of the role of information quality on consumer information adoption and decision-making behavior.From the perspective of online commentary utility,online product reviews,and online product reviews,the quality of online reviews is measured,and the impact of online review quality on consumer purchasing decisions is studied.In the research method,this paper chooses the multi-attribute decision-making model to sort the target goods.Therefore,in the model data processing,the total utility of online comments is first measured from the perspectives of online comment length,comment timeliness and picture comments.Then,the Apriori algorithm is used to extract the features of the online comment,and the online comment utility and product features are combined as the weight of the multi-attribute decision;the online comment emotion is divided into six levels by the sentiment dictionary,and the probability of falling into each emotional polarity is judged.As a judgment matrix for multi-attribute decision making.The target items are then sorted according to the TOPSIS algorithm.Finally,taking Tujia.com as an example,five randomly selected stores are used as target commodities,and the data are processed using the previously proposed models and algorithms to sort the five stores.In order to further verify the accuracy of the model and algorithm,the processing results of this thesis are compared with the methods proposed in the previous literature.Finally,the proposed algorithm and model have higher accuracy.The models and algorithms proposed in this paper can be applied to the following three scenarios.From the perspective of consumers,it is designed to help consumers choose better products in an objective sense and better assist consumers in making decisions.In this way,it can help merchants to better understand the needs of consumers.At the same time,from the perspective of the merchant platform,the research results of this paper can also help the platform to better sort similar products,making the sorting results more in line with the needs of consumers.