© 2019 IEEE. Recommendation systems play a key role in current online commerce enterprises. Despite their success, they usually behave like black-boxes from the user perspective, typically failing to produce high quality human-computer interactions; interpretability is thus a major concern for the next generation of recommendation systems. In this paper we propose a model-agnostic method based on topic models that generates explanations for content-based recommendation systems.
|Original language||American English|
|Number of pages||6|
|State||Published - 1 Oct 2019|
|Event||Proceedings - 2019 Brazilian Conference on Intelligent Systems, BRACIS 2019 - |
Duration: 1 Oct 2019 → …
|Conference||Proceedings - 2019 Brazilian Conference on Intelligent Systems, BRACIS 2019|
|Period||1/10/19 → …|