Semantic Representation Using Sub-Symbolic Knowledge in Commonsense Reasoning
The commonsense question and answering (CSQA) system predicts the right answer based on a comprehensive understanding of the question. Previous research has developed models that use QA pairs, the corresponding evidence, or the knowledge graph as an input. Each method executes QA tasks with represen...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-09-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/12/18/9202 |
| _version_ | 1797491469525712896 |
|---|---|
| author | Dongsuk Oh Jungwoo Lim Kinam Park Heuiseok Lim |
| author_facet | Dongsuk Oh Jungwoo Lim Kinam Park Heuiseok Lim |
| author_sort | Dongsuk Oh |
| collection | DOAJ |
| description | The commonsense question and answering (CSQA) system predicts the right answer based on a comprehensive understanding of the question. Previous research has developed models that use QA pairs, the corresponding evidence, or the knowledge graph as an input. Each method executes QA tasks with representations of pre-trained language models. However, the ability of the pre-trained language model to comprehend completely remains debatable. In this study, adversarial attack experiments were conducted on question-understanding. We examined the restrictions on the question-reasoning process of the pre-trained language model, and then demonstrated the need for models to use the logical structure of abstract meaning representations (AMRs). Additionally, the experimental results demonstrated that the method performed best when the AMR graph was extended with ConceptNet. With this extension, our proposed method outperformed the baseline in diverse commonsense-reasoning QA tasks. |
| first_indexed | 2024-03-10T00:48:58Z |
| format | Article |
| id | doaj.art-5b8ba7a6a05e4ac3926c06dc5fe6dd98 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-10T00:48:58Z |
| publishDate | 2022-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-5b8ba7a6a05e4ac3926c06dc5fe6dd982023-11-23T14:54:47ZengMDPI AGApplied Sciences2076-34172022-09-011218920210.3390/app12189202Semantic Representation Using Sub-Symbolic Knowledge in Commonsense ReasoningDongsuk Oh0Jungwoo Lim1Kinam Park2Heuiseok Lim3Department of Computer Science and Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, KoreaDepartment of Computer Science and Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, KoreaHuman-Inspired AI and Computing Research Center, Korea University, Seoul 02841, KoreaDepartment of Computer Science and Engineering, Korea University, 145 Anam-ro, Seongbuk-gu, Seoul 02841, KoreaThe commonsense question and answering (CSQA) system predicts the right answer based on a comprehensive understanding of the question. Previous research has developed models that use QA pairs, the corresponding evidence, or the knowledge graph as an input. Each method executes QA tasks with representations of pre-trained language models. However, the ability of the pre-trained language model to comprehend completely remains debatable. In this study, adversarial attack experiments were conducted on question-understanding. We examined the restrictions on the question-reasoning process of the pre-trained language model, and then demonstrated the need for models to use the logical structure of abstract meaning representations (AMRs). Additionally, the experimental results demonstrated that the method performed best when the AMR graph was extended with ConceptNet. With this extension, our proposed method outperformed the baseline in diverse commonsense-reasoning QA tasks.https://www.mdpi.com/2076-3417/12/18/9202abstract meaning representationsemantic representationsub-symboliccommonsense reasoningConceptNetcommonsense question and answering |
| spellingShingle | Dongsuk Oh Jungwoo Lim Kinam Park Heuiseok Lim Semantic Representation Using Sub-Symbolic Knowledge in Commonsense Reasoning Applied Sciences abstract meaning representation semantic representation sub-symbolic commonsense reasoning ConceptNet commonsense question and answering |
| title | Semantic Representation Using Sub-Symbolic Knowledge in Commonsense Reasoning |
| title_full | Semantic Representation Using Sub-Symbolic Knowledge in Commonsense Reasoning |
| title_fullStr | Semantic Representation Using Sub-Symbolic Knowledge in Commonsense Reasoning |
| title_full_unstemmed | Semantic Representation Using Sub-Symbolic Knowledge in Commonsense Reasoning |
| title_short | Semantic Representation Using Sub-Symbolic Knowledge in Commonsense Reasoning |
| title_sort | semantic representation using sub symbolic knowledge in commonsense reasoning |
| topic | abstract meaning representation semantic representation sub-symbolic commonsense reasoning ConceptNet commonsense question and answering |
| url | https://www.mdpi.com/2076-3417/12/18/9202 |
| work_keys_str_mv | AT dongsukoh semanticrepresentationusingsubsymbolicknowledgeincommonsensereasoning AT jungwoolim semanticrepresentationusingsubsymbolicknowledgeincommonsensereasoning AT kinampark semanticrepresentationusingsubsymbolicknowledgeincommonsensereasoning AT heuiseoklim semanticrepresentationusingsubsymbolicknowledgeincommonsensereasoning |