Supporting Movie Production: A Recommender Approach
Qiong Jia, Jing Li Zhou, Yajie Gu, Shouxun Liu, Jiguang Wu · 2018
In order to view the visual effect of scripts, movie and television creators (or directors) often resort to the Web for reference material in the process of movie and television production. These video materials are then used for editing and completing an entire story script. Existing retrieval systems and tools can hardly meet such needs. This is because simply relying on several keywords entered by users would not best model such complex user needs as the one involving a search for movie footage with related scenes or characters. To solve this problem, we make use of Semantic Web technologies to build a knowledge base of the movie domain. Through the semantic extension of user queries, the proposed recommender can reason about its knowledge base to obtain desirable movie and television materials for its user. We conducted a series of experiments for usability test. The results show that our recommender system has obvious advantages over the traditional keyword based query system in support of movie production.