Takuya Kitazawa is a product developer and data ethicist, working at the intersection of technological and societal aspects of data-driven applications. He professionally serves as a full-stack software & machine learning engineer, data scientist, and product manager, while advocating ethical product development as an OSS developer and technical evangelist.
Recommender system is a data-driven application that generates personalized content for users. This talk shows how Julia can be a deeply satisfying option to capture the unique characteristics of recommenders, which rely heavily on repetitive matrix computations in multi-stage data pipelines. To build trustworthy systems in terms of not only accuracy and scalability but usability and fairness at large, we particularly focus on API design and evaluation methods implemented on Recommendation.jl.