VideoTopic: Modeling User Interests for Content-Based Video Recommendation

VideoTopic: Modeling User Interests for Content-Based Video Recommendation

Qiusha Zhu (Department of Electrical and Computer Engineering, University of Miami, Coral Gables, FL, USA), Mei-Ling Shyu (Department of Electrical and Computer Engineering, University of Miami, Coral Gables, FL, USA) and Haohong Wang (TCL Research America, Santa Clara, CA, USA)
DOI: 10.4018/ijmdem.2014100101
OnDemand PDF Download:
List Price: $37.50
10% Discount:-$3.75


With the vast amount of video data uploaded to the Internet every day, how to analyze user interests and recommend videos that they are potentially interested in is a big challenge. Most video recommender systems limit the content to metadata associated with videos, which could lead to poor recommendation results since the metadata is not always available or correct. On the other side, visual content of videos contain information of different granularities, from a whole video, to portions of a video, and to an object in a video, which are not fully explored. This extra information is especially important for recommending new items when no user profile is available. In this paper, a novel recommendation framework, called VideoTopic, that targets at cold-start items is proposed. VideoTopic focuses on user interest modeling and decomposes the recommendation process into interest representation, interest discovery, and recommendation generation. It aims to model user interests by using a topic model to represent the interests in the videos and then discover user interests from user watch histories. A personalized list is generated to maximize the recommendation accuracy by finding the videos that most fit the user's interests under the constraints of some criteria. The optimal solution and a practical system of VideoTopic are presented. Experiments on a public benchmark data set demonstrate the promising results of VideoTopic.
Article Preview

1. Introduction

As the exponential wide use of digital devices, Internet video traffic will be 55 percent of all consumers’ Internet traffic in 2016, up from 51 percent in 2011, as reported by Cisco. It would take over 6 million years to watch the amount of videos that will cross global IP networks each month in 2016 (Cisco, 2012). Users are overloaded by the choices of so many videos that a smart recommender system is on demand, which could provide the recommendation lists personalized to each user’s interests.

Research on recommendations are generally proceeded along three dimensions: content-based recommendation that focuses on analyzing the content of items; collaborative filtering that utilizes user profiles, such as ratings or clicks, to recommend items for like-minded users; and hybrid recommendation that incorporates both approaches. Due to the superior performance in Netflix competition (Lowe, 1999), latent factor model (LFM) was adopted in many state-of-the-art recommendation models (Koren, Bell, & Volinsky, 2009; Rendle & Schmidt-Thieme, 2008; Salakhutdinov & Mnih, 2008). These approaches belong to the collaborative filtering category, which involve analyzing user profiles, typically in the form of the user-item matrix. However, in many situations, user profiles are not available or very sparse, especially for online videos, since a large proportion of the users browse videos anonymously. As a result, dealing with the cold-start problem is inevitable. The cold-start problem describes the scenarios in recommender systems when user profiles are not available, which commonly arises at the beginning of a recommender system. Thus, for cold-start items, i.e., items without any user behavior data, collaborative filtering would fail. Some recently proposed frameworks bring the contents of items into consideration, such as (Agarwal & Chen, 2009; Gantner, Drumond, Freudenthaler, Rendle, & Schmidt-Thieme, 2010), which extended LFM to incorporate item features, user features, and global features in its model. These approaches can only handle cold-start problem to some extent for they rely on factorizing the user-item matrix or use it to optimize the models. If all the items are cold-start items, which are very common in real applications, especially when launching a new recommender system, these improved approaches would still fail and only content-based recommendation can be adopted at this stage before enough user profiles can be gathered.

Complete Article List

Search this Journal:
Volume 14: 1 Issue (2023): Forthcoming, Available for Pre-Order
Volume 13: 4 Issues (2022): 1 Released, 3 Forthcoming
Volume 12: 4 Issues (2021)
Volume 11: 4 Issues (2020)
Volume 10: 4 Issues (2019)
Volume 9: 4 Issues (2018)
Volume 8: 4 Issues (2017)
Volume 7: 4 Issues (2016)
Volume 6: 4 Issues (2015)
Volume 5: 4 Issues (2014)
Volume 4: 4 Issues (2013)
Volume 3: 4 Issues (2012)
Volume 2: 4 Issues (2011)
Volume 1: 4 Issues (2010)
View Complete Journal Contents Listing