Papers with Code 아카이브

Story Completion

태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.

하위 태스크
0
평가표
0
보존된 평가 기록
0

Given a story prefix and two possible endings, determining which one is the correct (coherent) ending of the story.

데이터셋

출처에 같은 태스크 이름으로 연결된 데이터셋입니다. 동명 태스크가 포함될 수 있으므로, 비교할 때는 평가 결과의 조건도 확인하세요.

XStoryCloze

XStoryCloze consists of the professionally translated version of the English StoryCloze dataset (Spring 2016 version) to 10 non-English languages. This dataset is intended to be used for evaluating the zero- and few-shot learning capabilities of multlingual language models. This dataset is released by Meta AI.

Scifi TV Shows

A collection of long-running (80+ episodes) science fiction TV show synopses, scraped from Fandom.com wikis. Collected Nov 2017. Each episode is considered a "story". Contains plot summaries from : * Babylon 5 (https://babylon5.fandom.com/wiki/Main_Page) - 84 stories * Doctor Who (https://tardis.fandom.com/wiki/Doctor_Who_Wiki) - 311 stories * Doctor Who spin-offs - 95 stories * Farscape (https://farscape.fandom.com/wiki/Farscape_Encyclopedia_Project:Main_Page) - 90 stories * Fringe (https://fringe.fandom.com/wiki/FringeWiki) - 87 stories * Futurama (https://futurama.fandom.com/wiki/Futurama_Wiki) - 87 stories * Stargate (https://stargate.fandom.com/wiki/Stargate_Wiki) - 351 stories * Star Trek (https://memory-alpha.fandom.com/wiki/Star_Trek) - 701 stories * Star Wars books (https://starwars.fandom.com/wiki/Main_Page) - 205 stories * Star Wars Rebels - 65 stories * X-Files (https://x-files.fandom.com/wiki/Main_Page) - 200 stories Total: 2276 stories Dataset is "eventified" and generalized (see LJ Martin, P Ammanabrolu, X Wang, W Hancock, S Singh, B Harrison, and MO Riedl. Event Representations for Automated Story Generation with Deep Neural Nets, Thirty-Second AAAI Conference on Artificial Intelligence (AAAI), 2018. for details on these processes.) and split into train-test-validation sets for converting events into full sentences.

출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.