User interactions with online multimedia provide multiple aspects of users' dynamic preferences and behavior. My interest lies in understanding and modeling such dynamic preferences to build more effective music, movie or photo recommendations. Apart from analyzing user interactions, I am also exploring the importance of location in user behavior and establishing meaningful context for more diverse and novel recommendations.
"I like to explore sometimes": Adapting to Dynamic User Novelty Preferences
| ACM RecSys 2015| Paper| Slides
Jello can be found in different color and shape. Wise jello-ers believe that people like Jello often of only certain color and shape. However, based on my jello experience, I find that my friends do not always like the same color and shape. And that, though they prefer the green Jello shots in round cups, they do sometime prefer the Jello of different shape and color. As I believed, I saw similar dilemma among various other jello lovers. Thus, I build a Jello maker that can make the right color and shape that my friends are likely to try.
"Where Far Can Be Close": Finding Distant Neighbors In Recommendation Systems
| LocalRec, ACM RecSys 1st workshop on Location Aware Recommendations 2015| Paper| Slides
INDU: Making everyone more aware and responsible towards Global Warming
| 16th ACM International Workshop on Mobile Computing Systems and Applications, HotMobile, 2015| Paper
More Efficient Tagging Systems with Tag Seeding
| ASE (previosly IEEE) SocialCom 2014| Paper| Slides |
Specialization, Homophily, and Gender in a Social Curation Site: Findings from Pinterest.
| ACM CSCW 2014| Paper Slides
An Extended Approach to Personalize the Web Search by Measuring User Relevance.
| SWWS 2014| Paper Slides
For any question/feedback/inquiry about my work or any dataset involved please email to vikas [at] cs [dot] umn [dot] edu