+ All Categories
Home > Science > Improving IMDb Movie Recommendations with Interactive Settings and Filters

Improving IMDb Movie Recommendations with Interactive Settings and Filters

Date post: 08-Jun-2015
Category:
Upload: simon-dooms
View: 440 times
Download: 2 times
Share this document with a friend
Description:
Poster about the MovieBrain project as presented during the ACM RecSys 2014 conference in Foster City (CA, USA), Oct 8, 2014 by Simon Dooms.
Popular Tags:
1
WiCa, Wireless & Cable, www.wica.intec.ugent.be Gaston Crommenlaan 8 box 201, 9050 Ghent, Belgium Ghent University, Department of Information Technology Simon Dooms [email protected] Toon De Pessemier [email protected] Luc Martens [email protected] Improving IMDb Movie Recommendations with Interactive Settings and Filters MIDDLEWARE BACK END FRONT END Google Chrome extension Webserver API HPC infrastructure CONTROL RESULTS DATASET IMDb public ratings & MovieTweetings dataset Interactive settings Genre filters User-specific, weighted hybrid recommendation strategy Machine learning Online public experiment >150 days active >80 users Users enjoy the enhanced movie recommendation experience Users’ interaction behavior is unique MyMediaLite Try it out at themoviebrain.com MovieBrain Manual user involvement and user-specific strategies are the future. Background processes User Data Caching Logging API Request controller HPC
Transcript
Page 1: Improving IMDb Movie Recommendations with Interactive Settings and Filters

WiCa, Wireless & Cable, www.wica.intec.ugent.be Gaston Crommenlaan 8 box 201, 9050 Ghent, Belgium

Ghent University, Department of Information Technology

Simon Dooms [email protected]

Toon De Pessemier [email protected]

Luc Martens [email protected]

Improving IMDb Movie Recommendations

with Interactive Settings and Filters

MIDDLEWARE BACK END FRONT END

Google Chrome extension Webserver API HPC infrastructure

CONTROL RESULTS DATASET

IMDb public ratings

&

MovieTweetings

dataset

Interactive settings

Genre filters

User-specific, weighted

hybrid recommendation

strategy

Machine learning

Online public experiment

>150 days active

>80 users

Users enjoy the enhanced

movie recommendation

experience

Users’ interaction behavior

is unique

MyMediaLite

Try it out at themoviebrain.com

MovieBrain

Manual user

involvement and

user-specific strategies

are the future.

Background

processes

User Data

Caching

Logging

APIRequest

controllerHPC

Recommended