Through the establishment of commodity knowledge map, the associated information between commodities can be used to effectively recommend commodities, and the characteristic information of commodities in the map can also be used to assist customer service work.

Recommendation application of knowledge graph in e-commerce field

  • Knowledge graph is the attribute network of entities, which can well express the relationship between entities. The relationship can be entities with the same attribute or entity relationship with upper and lower levels.

  • For the recommendation system, the entities in the map are not only recommended contents, but also contain user information or labels. Therefore, the knowledge map provides a good network of recommended objects.

  • Through the knowledge graph, the recommendation system can recommend relevant content to you well. For example, if you buy a mobile phone, it can recommend charging treasure, protective cover, and toughened film to you, because it knows these products are accessories of the mobile phone in its mind.

  • 1. Recommendation of upper and lower related products

    The subordinate relation is the most relevant relation, which usually contains the meaning of the next operation and necessary conditions, for example: if you buy a mobile phone, you will buy a mobile phone case, if you buy a car, you will buy glass water, etc.

  • 2. Recommending products with combination attributes

    Some entities in the graph are connected by more than one parent entity, which usually has a strong correlation. For example, users search QC30 records and recommend QC35, QC25 and other results, which can not only improve the possibility of transaction, but also achieve a higher customer unit price, and achieve a win-win situation for merchants and platforms.

Application of knowledge graph in customer service in e-commerce field

  • Still learning, sorting…

The resources

  • Baijiahao.baidu.com/s?id=161028…
  • Baijiahao.baidu.com/s?id=160264…