‘Recommendation’ is something you come across every day on almost every online platform, from the morning news stories to the late-night online TV shows. The personalized suggestions “Guess you like” or “You might like” are given using AI-powered technology, predicting a user’s general behavior and preferences. But technology is developing day by day and enhancing its productivity. Thus, Google, being one of the leading companies in recommender system research, development, and deployment, has recently introduced TensorFlow Recommenders (TFRS), a new open-sourced TensorFlow package.
At Google, TensorFlow core has spent several years exploring deep learning techniques like multi-task learning, reinforcement learning, better user representations, and fairness objectives, to make more personalized and useful recommendations. Using all these, TensorFlow Recommenders makes building, evaluating, and serving sophisticated recommender models easy and efficient.
Features
Built with TensorFlow 2.x, features of TFRS are as follows:
- It helps to develop and evaluate flexible candidate nomination models
- It incorporates items, users, and context information into recommendation models easily
- It trains multi-task models that help optimize multiple recommendation objectives;
- It serves the final models using TensorFlow Serving.
These features are easy to use as TFRS contains different modules that help users customize individual layers and metrics easily. TFRS also forms a cohesive whole that allows the individual components to coordinate well. The emphasis is on making the default settings sensible, making common tasks intuitive and straightforward to implement, and providing more complex and custom recommendations, making it more flexible. Researchers also aim at making TFRS a better platform for conducting academic research.
Github: https://github.com/tensorflow/recommenders
TF Blog: https://blog.tensorflow.org/2020/09/introducing-tensorflow-recommenders.html?linkId=100309856
Shilpi is a Contributor to Marktechpost.com. She is currently pursuing her third year of B.Tech in computer science and engineering from IIT Bhubaneswar. She has a keen interest in exploring latest technologies. She likes to write about different domains and learn about their real life applications.
