Explainable recommendations — why opening black boxes matters

Original article was published on Artificial Intelligence on Medium

While explainable recommendations have seen increased attention in research over the last few years, few solutions have been deployed into production in industry. The next part of the series will cover my implementation of an explainable movie recommender system using two post-hoc explainers, it’s evaluation both offline and through a user study, as well as a discussion on how the explanations could be deployed into real systems.

The following references directly relate to the content of this post. A full reference list for the entire project is listed in my dissertation.

[1]: Koren, Y., & Bell, R. (2015). Advances in collaborative filtering. In Recommender systems handbook (pp. 77–118). Springer, Boston, MA.

[2]: Ricci, F., Rokach, L., & Shapira, B. (2011). Introduction to recommender systems handbook. In Recommender systems handbook (pp. 1–35). Springer, Boston, MA.

[3]: Koren, Y., Bell, R., & Volinsky, C. (2009). Matrix factorization techniques for recommender systems. Computer, (8), 30–37.

[4]: Tintarev, N., & Masthoff, J. (2011). Designing and evaluating explanations for recommender systems. In Recommender systems handbook (pp. 479–510). Springer, Boston, MA.

[5]: Sinha, R., Swearingen, K.: The role of transparency in recommender systems. In: Conference on Human Factors in Computing Systems, pp. 830–831 (2002)

[6]: Wang, X., Chen, Y., Yang, J., Wu, L., Wu, Z., & Xie, X. (2018, November). A Reinforcement Learning Framework for Explainable Recommendation. In 2018 IEEE International Conference on Data Mining (ICDM) (pp. 587–596). IEEE.

[7]: Zhang, Y., Lai, G., Zhang, M., Zhang, Y., Liu, Y., & Ma, S. (2014, July). Explicit factor models for explainable recommendation based on phrase-level sentiment analysis. In Proceedings of the 37th international ACM SIGIR conference on Research & development in information retrieval (pp. 83–92). ACM.

[8]: Peake, G., & Wang, J. (2018, July). Explanation mining: Post hoc interpretability of latent factor models for recommendation systems. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining (pp. 2060–2069). ACM.

[9]: Amazon.com. (2010). Improve Your Recommendations. Retrieved October 2, 2019, from https://www.amazon.com/gp/help/customer/display.html/ref=hp_16465201_FAQ_recommendations?nodeId=13316081.

[10]: Sethuraman, R. (2019, March 31). Why Am I Seeing This? We Have an Answer for You. Retrieved October 2, 2019, from https://newsroom.fb.com/news/2019/03/why-am-i-seeing-this/.