After understanding the word frequency analysis technology and application of product titles, operators can also apply this technology to the analysis of product review texts. We will take a clothing product listing with a high number of reviews as an example. Its product ASIN is B07G599PB7 and the listing title is “Amoretu Women Summer TunicDress V Neck Casual Loose Flowy Swing Shift Dresses”.
As of February 8, 2021, the listing link has 23,930 reviews and its review rating is 4 stars. After using the Python crawler script to crawl all the review texts, delete non-adjectives such as “a”, “the”, “I”, “to”, “do”, “is”, and “was”, and then save the frequency of occurrence of different adjectives in an Excel table. The author has saved this data in an Excel file named “Frequencies Of Keywords Reviewsi.” “NumOfWord” represents the total number of words, and the second column represents different adjectives and their frequencies of occurrence. As the total number of words increases, the frequencies of occurrence of different adjectives will also change accordingly. The above data file was analyzed using Python visualization technology to obtain a positive order dynamic arrangement diagram of the review text word frequency analysis. The positive order dynamic arrangement diagram shows the change in the frequency of different adjectives when the review text increases from less to more. As the number of review words increases, the adjectives with increasing frequency are the words that frequently appear in the latest review text (reviews generated by users in the most recent period of time), representing the recent subjective description of the product by consumers. The reverse order dynamic arrangement diagram shows the change in the frequency of different adjectives when the review text decreases from more to less. As the number of review words decreases, the adjectives with increasing frequency are the most popular reviews. (The reviews ranked at the top are also the reviews with the most likes) Frequently appearing words in the text represent the product selling points most recognized by consumers, as well as subjective descriptions related to the user experience.
The word cloud chart shows that the words that consumers have the greatest subjective feelings about the product are “cute”, “little”, “small”, “short” and other words. If the operator’s own products are similar to the products in the listing, then you can choose to add the words that appear frequently in the review word frequency analysis to the listing keywords and five-point description, thereby increasing the probability of search exposure and keyword matching.