Understanding stories with large-scale common sense
Story understanding systems need to be able to perform commonsense reasoning, specifically regarding characters' goals and their associated actions. Some efforts have been made to form large-scale commonsense knowledge bases, but integrating that knowledge into story understanding systems remai...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
CEUR-WS
2020
|
| Online Access: | https://hdl.handle.net/1721.1/128901 |
| _version_ | 1826217462606069760 |
|---|---|
| author | Williams, Bryan Michael Lieberman, Henry A Winston, Patrick H |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Williams, Bryan Michael Lieberman, Henry A Winston, Patrick H |
| author_sort | Williams, Bryan Michael |
| collection | MIT |
| description | Story understanding systems need to be able to perform commonsense reasoning, specifically regarding characters' goals and their associated actions. Some efforts have been made to form large-scale commonsense knowledge bases, but integrating that knowledge into story understanding systems remains a challenge. We have implemented the Aspire system, an application of large-scale commonsense knowledge to story understanding. Aspire extends Genesis, a rule-based story understanding system, with tens of thousands of goalrelated assertions from the commonsense semantic network ConceptNet. Aspire uses ConceptNet's knowledge to infer plausible implicit character goals and story causal connections at a scale unprecedented in the space of story understanding. Genesis's rule-based inference enables precise story analysis, while ConceptNet's relatively inexact but widely applicable knowledge provides a significant breadth of coverage difficult to achieve solely using rules. Genesis uses Aspire's inferences to answer questions about stories, and these answers were found to be plausible in a small study. Though we focus on Genesis and ConceptNet, demonstrating the value of supplementing precise reasoning systems with large-scale, scruffy commonsense knowledge is our primary contribution. |
| first_indexed | 2024-09-23T17:04:03Z |
| format | Article |
| id | mit-1721.1/128901 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T17:04:03Z |
| publishDate | 2020 |
| publisher | CEUR-WS |
| record_format | dspace |
| spelling | mit-1721.1/1289012022-10-03T10:10:37Z Understanding stories with large-scale common sense Williams, Bryan Michael Lieberman, Henry A Winston, Patrick H Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Story understanding systems need to be able to perform commonsense reasoning, specifically regarding characters' goals and their associated actions. Some efforts have been made to form large-scale commonsense knowledge bases, but integrating that knowledge into story understanding systems remains a challenge. We have implemented the Aspire system, an application of large-scale commonsense knowledge to story understanding. Aspire extends Genesis, a rule-based story understanding system, with tens of thousands of goalrelated assertions from the commonsense semantic network ConceptNet. Aspire uses ConceptNet's knowledge to infer plausible implicit character goals and story causal connections at a scale unprecedented in the space of story understanding. Genesis's rule-based inference enables precise story analysis, while ConceptNet's relatively inexact but widely applicable knowledge provides a significant breadth of coverage difficult to achieve solely using rules. Genesis uses Aspire's inferences to answer questions about stories, and these answers were found to be plausible in a small study. Though we focus on Genesis and ConceptNet, demonstrating the value of supplementing precise reasoning systems with large-scale, scruffy commonsense knowledge is our primary contribution. Air Force Office of Scientific Research (Award FA9550-17-1-0081) 2020-12-22T20:58:35Z 2020-12-22T20:58:35Z 2017 2019-07-08T16:57:30Z Article http://purl.org/eprint/type/ConferencePaper https://hdl.handle.net/1721.1/128901 Williams, Bryan et al. "Understanding stories with large-scale common sense." Proceedings of the Thirteenth International Symposium on Commonsense Reasoning, November 2017, London, United Kingdom, CEUR-WS, 2017 © 2017 Association for the Advancement of Artificial Intelligence en https://dblp.org/db/conf/commonsense/index.html Proceedings of the Thirteenth International Symposium on Commonsense Reasoning Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf CEUR-WS MIT web domain |
| spellingShingle | Williams, Bryan Michael Lieberman, Henry A Winston, Patrick H Understanding stories with large-scale common sense |
| title | Understanding stories with large-scale common sense |
| title_full | Understanding stories with large-scale common sense |
| title_fullStr | Understanding stories with large-scale common sense |
| title_full_unstemmed | Understanding stories with large-scale common sense |
| title_short | Understanding stories with large-scale common sense |
| title_sort | understanding stories with large scale common sense |
| url | https://hdl.handle.net/1721.1/128901 |
| work_keys_str_mv | AT williamsbryanmichael understandingstorieswithlargescalecommonsense AT liebermanhenrya understandingstorieswithlargescalecommonsense AT winstonpatrickh understandingstorieswithlargescalecommonsense |