Research On Title Optimization Based On Online Reviews

Since the birth of e-commerce, product title optimization has been a hot topic in practice. With the problem of homogeneity of products becoming more and more serious on online shopping platforms, consumers are always faced with many products that are similar in price, seller’s reputation, picture of the product, and so on. At this time, titles become more important. The essence of title optimization is the design of keywords in the titles. In order to improve the page view of their shops, retailers try their best to put the so-called “hot words” in the titles. However, on the e-commerce platforms, such as Taobao.com, the products’ titles are used not only for being searched by the search engine, but also for attracting consumers to click in.In this study, we make full use of online reviews and some theoretical achievements in marketing, informatics and bibliometrics to design a new framework to optimize the titles of products, especially for experience goods. This framework includes data collection,thesaurus compilation, feature extraction and title design. Considering the specificity of e-commerce when compiling the thesaurus, we add some technical terms into existing general vocabulary system as a supplement. A user experiment based on the reviews of women clothes, women shoes, digital products and office appliances is then carried out to test the validity of this method. The results show that, the optimized titles are significantly better than the original titles in click rate for the experience goods, while the search goods are not the case. Later we explain this phenomenon on the strength of the classical S-E-C framework to identify the difference in product attributes and consumer cognition.The quality of the title is directly related to the state of operation of a shop. And the problem of title optimization is even more important for those small and medium-sized online retailers lacking brand effect and scale effect. This new method can be flexibly utilized by those small and media online retailer who are not professional in this academic area because the dataset of online reviews is easy to get and the procedure of optimization is not that complicated. If applied properly, it will be sure to help retailers get more real flow of consumers.