Network Comments Data Mining-Based Analysis Method of Consumer’s Perceived Value

Abstract

The user comments on e-commerce websites convey what users think of the goods sold. Thus when deciding the price of a goods, sellers should take many elements into consideration, including cost, price given by the competitors, estimated profit and, what’s more, consumer’s perceived value. Based on the analysis of the factors which influence consumer’s perceived value, we introduces a term called consumer cognitive property of product for the first time. Meanwhile, a data mining method based on text clustering is proposed. We crawled millions of comments from e-commerce website. Then extracted the consumer cognitive properties by a variance cumulative method and cluster these properties based on K-means. After these steps, most impressive influencing factors of the product on consumer’s perceived value are extracted to assist the sellers to make final pricing strategies. An evaluation performed with the collections crawled from taobao shows that our algorithm works great.