Recommendation reason generation method and device

Abstract

The invention discloses a recommendation reason generation method and device. The method comprises the steps that comment content of each object to be recommended is excavated; aiming at the comment content of each object to be recommended, a classification attribute of a preset dimension is established, comment classification prediction is conducted, and a comment statement which belongs to eachclass of each object to be recommended is obtained; according to the comment statement, belonging to each class, of each object to be recommended, a recommendation reason of the object to be recommended is generated. By adopting the recommendation reason generation method, the comment content which is excavated from the comment content of the object to be recommended is used as the candidate recommendation reason of the object to be recommended, and the diversity of the recommendation reason is improved; according to the classification attribute of the preset dimension, the comment content, belonging to each class, of the object to be recommended is selected as the recommendation reason, then the pertinence of the recommendation reason is improved, and the effects are achieved that a useris attracted and the user conducts clicking or ordering motion on the object to be recommended.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a method and device for generating recommendation reasons. Background technique [0002] With the development of online and offline e-commerce (Online To Offline, referred to as O2O) transaction mode, how to make users have enough interest in the stores provided on the O2O e-commerce trading platform, and guide users to click or place an order to become an e-commerce business? Important issues for platform development. In the prior art, there are two main ways to solve the above problems: Method 1, the sales volume of each store is counted according to the city area and store category, and the ranking of the sales list is used as the reason for recommendation. For example, the reason for recommendation is Xihu District No. 1 hot pot; method 2, use the user's historical browsing footprint and transaction behavior records as the recommendation reason, for example, the re...

AI Technical Summary

Problems solved by technology

[0003] However, in the process of implementing the present invention, the inventor found at least the following problems in the prior art: the first method is to recommend according to the list, so that each store is a fixed type of list, that is: urban area + category list It is difficult to expand to other types of lists, so that all users see the same recommendation reasons, and for stores with similar distances and the same purpose, the recommendation reasons displayed will be homogeneous, that is, in the store On the list page, several stores in a row have the same recommendation reason; method 2 recommends according to user behavior. Although this method has a strong sense of identity with users and is a personalized display, for individuals, users click and trade. Stores that have been visited are only a small part, and most of the stores have no behavior, resulting in most stores not displaying the reason for recommendation, so the coverage rate is low

Images