Choudhary, Priyanka and Desarkar, Maunendra Sankar
(2018)
PReFacTO: Preference Relations Based
Factor Model with Topic Awareness and
Offset.
Masters thesis, Indian Institute of Technology Hyderabad.
Abstract
Recommendation systems create personalized list of items that might interest the user
by analyzing the user’s history of past purchases and/or consumption. Generally only
a small subset of the items are assessed by each user, and from the large subset of
unseen items, the systems need to produce an accurate list of recommendations.
For rating based systems, most of the traditional methods for recommendation
focus on the absolute ratings provided by the users to the items. In this work,
we extend the traditional Matrix Factorization approach for recommendation and
propose pairwise relation based factor modeling. We propose the method based on
the pairwise preferences between the items to capture the relative tendency of user
selecting one item over the other.
While modeling the items in the system, the use of pairwise preferences allow
information flow between the items through the preference relations as an additional
information. Item feedbacks are available in the form of reviews apart from the
rating information. The reviews have textual information that can be really helpful
to represent the item’s latent feature vector appropriately. We perform topic modeling
of the item reviews and use the topic vectors to guide the joint factor modeling of the
users and items and learn their final representations. The proposed methods shows
promising results in comparison to the state-of-the-art methods in our experiments.
vi
Actions (login required)
 |
View Item |
Altmetric
Altmetric