Knowledge graph-based precise-recommendation technology of offline medicine retail

Abstract

The invention provides knowledge graph-based precise-recommendation technology of offline medicine retail. The technology is based on a medicine knowledge graph. The graph mainly includes entities ofdiseases, Chinese and Western medicine, people groups, health products and the like and association relationships among the entities. The association relationships mainly include contraindication relationships, interaction relationships, inclusion relationships, adjoining relationships, assistance relationships, affiliation relationships and the like. A people group set is generated through a people group label generation algorithm based on density clustering, and contains people group category labels and medicine commodities belonging to all the labels; then a target user is labeled accordingto a medicine combination bought by the user; and finally, interest degrees of the user are calculated according to a label, and a recommendation result list is generated. The technology can solve the problems which are of sparseness and single categories of user data and are encountered by medicine retail enterprises in precise recommendation, and the medicine knowledge graph on which the technology is based can also solve the problem of particularity of the medicine commodities and the problem of medicine field knowledge lacking of users. The technology can satisfy most precise-recommendation service demand in an offline medicine retail environment, and has very high market competitiveness and broad application prospects.

Description

technical field [0001] The invention relates to the field of computer application technology, and relates to an accurate recommendation technology for offline pharmaceutical retailing based on knowledge graphs. Background technique [0002] At present, the competition among enterprises in the pharmaceutical retail industry is very fierce. Large-scale offline drugstores rapidly expand the scale of enterprises by increasing the number of stores to gain a competitive position in the market. Medicines and health products are closely related to health, and pharmaceutical sales personnel need to have professional knowledge related to medicine. , most chain drugstores are facing the problem of shortage of professional talents, and pharmaceutical retail enterprises urgently need to develop new ways to enhance their competitiveness. [0003] Accurate recommendation technology can help enterprises to mine users' interests in massive data, recommend products that users like, stimulate ...

AI Technical Summary

Problems solved by technology

[0005] There are many problems in the application process of traditional precision recommendation technology in the special background of offline pharmaceutical retail, including: the particularity of pharmaceutical products, the lack of knowledge in the medical field of users, the sparseness of historical drug purchase behavior data of some users, the problem of most users The problem of sparse purchase data of historical health care products, single type of user behavior data (all purchase behavior data)

[0006] Most of the current precise recommendations are for movies, music, news, daily consumer goods and other fields, and precise recommendation technologies that can solve the above-mentioned problems for offline pharmaceutical retail scenarios are rare

Images