Research On Finding Hot Commodities In Social Network

With the development of the network and the improvement of social logistics and finance,online shopping has shown an extremely rapid development trend in recent years.People in daily life are constantly in contact with a large amount of commodity information,which affects people's shopping choices in the subconscious.Due to the rapid rise of social networks such as Weibo,a large number of user generations information often lead the trend of the society,causing the goods to explode.This has led to the"Matthew effect"in Internet shopping.The higher the popularity,the higher the sales volume.When users select products,popular products mean that they are more acceptable to users.So how to extract product information from the unstructured network,and discover the popularity of the products,will provide reference for consumers'shopping choices,and also for merchants.Give opinions on production and sales decisions.The structure of the social network itself is not clear,and the content form is arbitrary and the colloquial problem is serious,which makes it difficult to obtain popular goods in social networks.This paper starts with how to obtain commodity information in social networks and how to obtain the popularity of extracted products.This paper proposes a hot commodity shopping guide algorithm for social networks.This paper first analyzes the definition of commodity entities in Weibo,and builds a physical library in the mobile phone field.An improved named entity recognition method is proposed through known domain knowledge.Through the product entity library,better product and feature recognition effects are achieved without segmentation of the corpus.Secondly,based on the entity identification,the product entity is standardized by the microblog propagation chain,and based on the improved weighted LeaderRank algorithm,the influential microblog text nodes in the social network are discovered,and the time decay factor is introduced.The product entity is standardized,and a discovery algorithm for hot commodities is proposed,and the feature recommendation of the product is given by matching the features.The experiment proves that the proposed method can identify the commodity information in the social network content well,and the hot items given have high user coverage.At the same time,this paper gives a shopping guide platform for hot commodities,which provides users and merchants with reference to commodity popularity information.