The entertainment industry is fraught with risks. Production companies must rigorously decide which planned entertainment products to turn into formal filming. However, information is limited before production. Given this constraint, this paper develops a comparable title identification method named ICTPP to identify past products comparable to the proposed entertainment product so that its potential demand can be accurately forecasted. The method first constructs a heterogeneous entertainment product network that simultaneously accommodates entertainment product nodes and three types of factor nodes available in the pre-production stage. Then, a meta-path-based network embedding method is proposed to learn entertainment product representations that preserve the latent structural and semantic information in the network. With experiments conducted on real-world viewership record data, we validated ICTPP's effectiveness and superiority in identifying comparable titles compared with existing studies.


Paper Number 1762; Track Design; Short Paper


When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.