Motivation
Spoilers are information that spoils the enjoyment of a TV program or movie, such as the final outcome of a suspense drama or movie, the end results of reality television shows and sports broadcasts. They are often transmitted over the Internet or through social networks such as Facebook and Twitter, where people share their opinions with friends or receive messages from strangers.
Researchers at the University of California San 먹튀검증소 Diego have created an artificial intelligence agent that filters out spoilers in SNS messages and reviews. The agent is trained by assembling a database of over a million spoiler-tagged reviews from Amazon-owned reading community Goodreads, then it uses machine learning to spot spoilers in sentences from that data and elsewhere. It’s able to label spoilers up to 92 percent of the time. It’s not perfect, and its accuracy depends on the data and model used. Nevertheless, it’s a step forward in the quest to automatically spot spoilers. Thanks to Mengting Wan, Rishabh Misra and Ndapa Nakashole for their paper, “Fine-Grained Spoiler Detection from Large-Scale Review Corpora,” in Proceedings of the Association for Computational Linguistics (ACL), pages 2605-2610.
0 comments:
Post a Comment