Published December 14, 2020
| Version v1
Conference paper
Injection of Knowledge in a Sourcing Recommender System
Description
Recommender systems provide suggestions to users for items that best meet their needs. In this work, we study the benefits of using knowledge and, more specifically, a 'bag of concepts' representation to enhance a recommender system in the sourcing domain. We tested our approach in a real-world case study provided by the Silex company. The experimental results show that injecting knowledge in the recommendation process outperforms word embedding approaches.
Abstract
International audience
Additional details
- URL
- https://hal.inria.fr/hal-02996442
- URN
- urn:oai:HAL:hal-02996442v1
- Origin repository
- UNICA