TY - GEN
T1 - Correcting Ambiguous Japanese Sentences Based on Japanese Dependency Analysis
AU - Nemoto, Shota
AU - Kamioka, Eiji
AU - Kanamaru, Manami
AU - Phan Xuan, Tan
N1 - Publisher Copyright:
© 2022 IEEE.
PY - 2022
Y1 - 2022
N2 - The purpose of this study is to correct ambiguous Japanese sentences into clear and understandable ones. Ambiguous sentences are ones that can be interpreted differently by different readers. To achieve this purpose, existing studies have corrected ambiguous sentences by replacing word order and inserting commas using word occurrence probabilistic models. However, such probabilistic models often produce sentences that do not reflect the writer's intention because they prioritize frequently used sentence patterns over sentence meanings. As a result, the corrected sentences can be clearly written but missing points. To solve the problem, this study proposes a new method which replaces the word order in a sentence focusing on the semantic connection between words and inserts commas, resulting in producing clear sentences without losing the writer's intention.
AB - The purpose of this study is to correct ambiguous Japanese sentences into clear and understandable ones. Ambiguous sentences are ones that can be interpreted differently by different readers. To achieve this purpose, existing studies have corrected ambiguous sentences by replacing word order and inserting commas using word occurrence probabilistic models. However, such probabilistic models often produce sentences that do not reflect the writer's intention because they prioritize frequently used sentence patterns over sentence meanings. As a result, the corrected sentences can be clearly written but missing points. To solve the problem, this study proposes a new method which replaces the word order in a sentence focusing on the semantic connection between words and inserts commas, resulting in producing clear sentences without losing the writer's intention.
KW - Japanese
KW - an ambiguous sentence
KW - comma
KW - word order
UR - http://www.scopus.com/inward/record.url?scp=85141436501&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85141436501&partnerID=8YFLogxK
U2 - 10.1109/AiIC54368.2022.9914578
DO - 10.1109/AiIC54368.2022.9914578
M3 - Conference contribution
AN - SCOPUS:85141436501
T3 - Proceedings - AiIC 2022: 2022 Applied Informatics International Conference: Digital Innovation in Applied Informatics during the Pandemic
SP - 13
EP - 18
BT - Proceedings - AiIC 2022
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2022 Applied Informatics International Conference, AiIC 2022
Y2 - 18 May 2022 through 19 May 2022
ER -