Investigating and Enhancing The Attentive Item2vec (AI2V) Model for Collaborative Filtering

Keren Gaiger,

M.Sc. student at the department of Industrial Engineering in Tel Aviv University

31 March 2022, 12:30 PM, Room 206& via zoom

Abstract:
Factorization methods for recommender systems tend to represent users as a single latent vector. However, user behavior and interests may change in the context of the recommendations that are presented to the user. For example, in the case of movie recommendations, it is usually true 
that earlier user data is less informative than more recent data. However, it is possible that a certain early movie may become suddenly more relevant in the presence of a popular sequel movie. This is just a single example of a variety of possible dynamically altering user interests in the pres- 
ence of a potential new recommendation. We propose the Attentive Item2vec (AI2V) which employs a context-target attention mechanism in order to learn and capture different characteristics of user historical behavior 
(context) with respect to a potential recommended item (target). We present extensive quantitative evaluations on five datasets, where it is shown that AI2V outperforms other baselines over all datasets and perform qualitative evaluations and demonstrate how AI2V can be beneficial for explainability on recommendations presented to the user. 

Bio:
Keren Gaiger, M.Sc. student at the department of Industrial Engineering in Tel Aviv University, specializing in Data science. Keren holds a B.Sc. degree in Industrial Engineering from the Technion. At her research she implemented a recommender system using an Attention mechanism taken from the  NLP domain. The research is being supervised by Dr. Noam Keonigstein