Recent Publications

Publications The latest 10 papers published or under review

Mitigating Stereotypes in Word Embedding through Sentiment Modulation

Huije Lee, Jin-Woo Chung, and Jong C. Park
Korea Software Congress (KSC), Pyeongchang, Korea, December 19-21, 2018.
(Accepted)

Neural Grammatical Error Correction by Simulating the Human Learner and the Human Proofreader

Fitsum Gaim, Jin-Woo Chung, and Jong C. Park
Korea Software Congress (KSC), Pyeongchang, Korea, December 19-21, 2018.
(Accepted)

Feature Attention Network: Interpretable Depression Detection from Social Media

Hoyun Song, Jinseon You, Jin-Woo Chung, and Jong C. Park
32nd Pacific Asia Conference on Language, Information and Computation (PACLIC 32), The Hong Kong Polytechnic University, Hong Kong SAR, December 1-3, 2018.
(Accepted)

Extracting Supporting Evidence with High Precision via Bi-LSTM Network

ChaeHun Park, Wonsuk Yang, and Jong C. Park
30th Annual Conference on Human & Cognitive Language Technology, Korea University, Seoul, Korea, October 12-13, 2018.
Show abstract
끉吏媛 넂 꽕뱷젰쓣 媛뽮린 쐞빐꽌뒗 異⑸텇븳 吏吏 洹쇨굅媛 븘슂븯떎. 끉吏 궡쓽 二쇱옣쓣 끉由ъ쟻쑝濡 吏吏븷 닔 엳뒗 洹쇨굅 옄猷 異붿텧쓽 옄룞솕뒗 옄룞 넗濡 떆뒪뀥, 젙梨 닾몴뿉 븳 쓽궗 寃곗젙 蹂댁“ 벑 뿬윭 뼱뵆由ъ씠뀡쓽 媛쒕컻 諛 긽슜솕瑜 쐞빐 븘닔쟻쑝濡 빐寃곕릺뼱빞 븳떎. 븯吏留 쎒臾몄꽌濡쒕꽣 吏吏 洹쇨굅瑜 異붿텧븯뒗 떆뒪뀥쓣 쐞빐꽌뒗 떎쓬怨 媛숈 몢 媛吏 뿰援ш 꽑뻾릺뼱빞 븯怨, 씠뒗 넂 꽦뒫쓽 떆뒪뀥 援ы쁽쓣 뼱졄寃 븳떎: 1) 끉吏쓽 二쇱젣 吏곸젒쟻씤 愿젴꽦 궙吏留 吏吏 洹쇨굅濡 궗슜맆 닔 엳뒗 젙蹂대 솗蹂댄븯湲 쐞븳 꼻 寃깋 踰붿쐞, 2) 닔吏묓븳 젙蹂 궡뿉꽌 끉吏쓽 二쇱옣쓣 紐낇솗븯寃 吏吏븷 닔 엳뒗 洹쇨굅瑜 떇蹂꾪븷 닔 엳뒗 씤吏 뒫젰. 蹂 뿰援щ뒗 넂 젙諛룄 솗옣 媛뒫꽦쓣 媛吏 吏吏 洹쇨굅 異붿텧쓣 쐞빐 떎쓬怨 媛숈 떒怨꾩쟻 吏吏 洹쇨굅 異붿텧 떆뒪뀥쓣 젣븞븳떎: 1) TF-IDF 쑀궗룄 湲곕컲 愿젴 臾몄꽌 꽑蹂, 2) 쓽誘몄쟻 쑀궗룄瑜 넻븳 吏吏 洹쇨굅 1李 異붿텧, 3) 떊寃쎈쭩 遺꾨쪟湲곕 넻븳 吏吏 洹쇨굅 2李 異붿텧. 젣븞븯뒗 떆뒪뀥쓽 쑀슚꽦쓣 寃利앺븯湲 쐞빐 궗꽕 4008媛 궡쓽 二쇱옣뿉 빐 쎒 긽뿉 엳뒗 845675媛쒖쓽 돱뒪뿉꽌 吏吏 洹쇨굅瑜 異붿텧븯뒗 떎뿕쓣 닔뻾븯떎. 二쇱옣怨 吏吏 洹쇨굅瑜 二쇱꽍븳 젙蹂댁뿉 븯뿬 꽦뒫 룊媛瑜 吏꾪뻾븳 寃곌낵 蹂 뿰援ъ뿉꽌 젣븞븳 떒怨꾩쟻 떆뒪뀥 1,2李 異붿텧 怨쇱젙뿉꽌 媛곴컖 0.41, 0.70쓽 젙諛룄瑜 蹂댁떎. 씠썑 떆뒪뀥씠 異붿텧븳 吏吏 洹쇨굅瑜 遺꾩꽍븯뿬, 끉吏뿉 븳 쟻젅븳 씠빐瑜 諛뷀깢쑝濡 븳 吏吏 洹쇨굅 異붿텧씠 媛뒫븯떎뒗 寃껋쓣 솗씤븯떎.

Automatic Tension Recognition from Lecture Show Transcripts

Seungwon Yoon, Wonsuk Yang, and Jong C. Park
30th Annual Conference on Human & Cognitive Language Technology, Korea University, Seoul, Korea, October 12-13, 2018.
Show abstract
湲댁옣씠씪뒗 痢〓㈃ 쓽궗냼넻쓣 븯嫄곕굹 湲쓣 씫쓣 븣 궗엺뿉寃 빆긽 쁺뼢쓣 二쇨퀬 엳떎. 湲댁옣쓽 媛쒕뀗 옄뿰뼵뼱泥섎━ 遺꾩빞뿉꽌 愿묐쾾쐞븳 쓽誘몃줈 궗슜릺뼱 솕뒗뜲, 蹂 끉臾몄 씠윴 媛쒕뀗 以 媛뺤뿰怨 媛숈 븳 諛⑺뼢 솕뿉꽌 솕옄쓽 留먯뿉 븯뿬 泥以묒씠 媛吏뒗 湲댁옣룄뿉 吏묒쨷븯뿬 씠瑜 젙웾솕븯뒗 諛⑸쾿쓣 젣븞븳떎. 븳 紐낆쓽 옄뿉 쓽빐 꽌닠맂 臾몄꽌뿉 湲댁옣룄 媛쒕뀗쓣 쟻슜븿뿉 엳뼱, 븳 諛⑺뼢 솕뿉꽌쓽 湲댁옣룄瑜 젙웾솕븯뒗 蹂 뿰援щ뒗 湲댁옣룄 媛쒕뀗쓣 씪諛 臾몄꽌뿉 쟻슜븷 븣뿉 蹂대떎 슜씠븯寃 솢슜맆 寃껋쑝濡 삁긽븳떎. 蹂 뿰援ъ뿉꽌뒗 癒쇱 솕옄쓽 留먯뿉 븳 泥以묒쓽 湲댁옣룄媛 二쇱꽍릺뼱 엳뒗 깉濡쒖슫 留먮춬移섎 援ъ텞븯떎. 삉븳 臾몃㎘쓣 怨좊젮븯뿬 湲댁옣룄瑜 삁痢≫븷 닔 엳뒗 紐⑤뜽怨 씠뿉 뵲瑜 湲댁옣룄 遺꾨쪟 꽦뒫뿉 븳 떎뿕 寃곌낵瑜 넻븯뿬 옄룞 湲댁옣룄 遺꾨쪟媛 怨꾩궛쟻쑝濡 媛뒫븯떎뒗 寃껋쓣 蹂댁씤떎.

Extracting Spatial Information about Events from Text

Jin-Woo Chung
PhD Dissertation, KAIST, Feb. 2018

Detection of Non-Standard Meaning Usage with Word Embedding

Huije Lee, Hancheol Park, Wonsuk Yang, and Jong C. Park
Human-Computer Interaction Korea (HCI), Jeongseon, Korea, January 31-February 2, 2018.
Show abstract
蹂 뿰援ъ뿉꽌뒗 遺꾩궛 몴긽 湲곕쾿쑝濡 뀓뒪듃뿉꽌 궗쟾긽쓽 쓽誘몃줈 궗슜릺吏 븡 뼱쐶(씠븯, 鍮꾪몴以 쓽誘 뼱쐶)瑜 깘吏븯뒗 紐⑤뜽쓣 젣븞븳떎. 뼱쐶쓽 뼱삎 룞씪븯굹 鍮꾪몴以 쓽誘몃줈 궗슜릺뒗 寃쎌슦瑜 뙋떒븯뒗 寃껋 옄룞솕맂 뀓뒪듃 遺꾩꽍 諛 삤뿭쓽 臾몄젣瑜 빐寃고븯뒗 뜲 以묒슂븳 슂냼씠떎. 蹂 뿰援ъ뿉꽌뒗 遺꾩궛 몴긽 湲곕쾿쑝濡 깮꽦맂 臾몃㎘ 諛 긽 떒뼱 踰≫꽣瑜 씠슜븯뿬, 긽 떒뼱媛 二쇱뼱吏 臾몃㎘ 궡뿉꽌 쟻빀븳吏瑜 寃利앺븯怨 긽 떒뼱媛 鍮꾪몴以 쓽誘몃줈 궗슜릺뿀뒗吏 뿬遺瑜 뙋떒븳떎. 蹂 뿰援ъ뿉꽌뒗 湲곗〈 뿰援ъ뿉꽌쓽 臾몃㎘ 踰≫꽣 깮꽦 諛⑹떇씠 吏땲뒗 臾몄젣젏쓣 빐寃고븯湲 쐞빐, 넻빀쟻씤 臾몃㎘ 젙蹂대 몴긽븯뒗 諛⑸쾿怨 臾몃㎘ 궡 떒뼱뱾쓽 媛以묒튂瑜 二쇰뒗 諛⑸쾿쓣 젣븞븳떎. 젣븞븯뒗 諛⑸쾿 듃쐞꽣 뜲씠꽣瑜 씠슜븳 떎뿕뿉꽌 湲곗〈뿉 젣븞맂 紐⑤뜽蹂대떎 뜑 넂 꽦뒫쓣 蹂댁떎.

Predicting Symptoms of Depression for Social Media Users via Linguistic Patterns

Hoyun Song, Hancheol Park, Wonsuk Yang, and Jong C. Park
Korea Software Congress (KSC), Busan, Korea, December 20-22, 2017.
Show abstract
슦슱利앹 媛쒖씤쓽 씪긽 湲곕뒫 븯 諛 떎뼇븳 궗쉶쟻 臾몄젣瑜 빞湲고븷 닔 엳湲 븣臾몄뿉 議곌린 吏꾨떒씠 以묒슂븯떎. 씠윭븳 議곌린 吏꾨떒쓽 떆룄濡쒖꽌, 蹂 뿰援щ뒗 냼뀥 誘몃뵒뼱 뀓뒪듃瑜 씠슜븯뿬 궗슜옄뱾쓽 슦슱利 뿬遺瑜 삁痢≫븯뒗 紐⑤뜽쓣 젣븞븳떎. 蹂 뿰援ъ뿉꽌뒗 鍮꾩젙삎 뀓뒪듃씤 냼뀥 誘몃뵒뼱 뀓뒪듃 긽뿉꽌 湲곗〈쓽 뼱쐶 湲곕컲 紐⑤뜽씠 吏땶 븳怨꾩젏씤 뼱쐶 留ㅼ묶 臾몄젣 諛 슦슱利앹쓣 寃り퀬 엳吏 븡 궗슜옄뱾쓽 슦슱利 愿젴 뼱쐶 궗슜怨 愿젴븳 臾몄젣젏쓣 빐寃고븯湲 쐞빐, 蹂대떎 떖痢듭쟻씤 뼵뼱븰쟻 뙣꽩쓣 씠슜븳 紐⑤뜽쓣 젣떆븳떎. 蹂 뿰援ъ쓽 떎뿕쓣 넻빐 궗슜옄쓽 슦슱利 뿬遺瑜 삁痢≫븿뿉 엳뼱 뼵뼱븰쟻 뙣꽩쓣 븿猿 쟻슜븷 寃쎌슦 떒닚븳 뼱쐶 湲곕컲 紐⑤뜽뿉 鍮꾪빐 뜑슧 슚怨쇱쟻엫쓣 솗씤븷 닔 엳뿀떎.

Extraction of Gene-Environment Interaction from the Biomedical Literature

Jinseon You, Jin-Woo Chung, Wonsuk Yang, and Jong C. Park
Proceedings of the 8th International Joint Conference on Natural Language Processing (IJCNLP 2017), pp. 865874, Taipei, Taiwan, November 27밆ecember 1, 2017.
Show abstract
Genetic information in the literature has been extensively looked into for the purpose of discovering the etiology of a disease. As the gene-disease relation is sensitive to external factors, their identification is important to study a disease. Environmental influences, which are usually called Gene-Environment interaction (GxE), have been considered as important factors and have extensively been researched in biology. Nevertheless, there is still a lack of systems for automatic GxE extraction from the biomedical literature due to new challenges: (1) there are no preprocessing tools and corpora for GxE, (2) expressions of GxE are often quite implicit, and (3) document-level comprehension is usually required. We propose to overcome these challenges with neural network models and show that a modified sequence-to-sequence model with a static RNN decoder produces a good performance in GxE recognition.

Inferring Implicit Event Locations from Context with Distributional Similarities

Jin-Woo Chung, Wonsuk Yang, Jinseon You, and Jong C. Park
Proceedings of the 26th International Joint Conference on Artificial Intelligence (IJCAI-17), pp. 979-985, Melbourne, Australia, August 19-25, 2017.
Show abstract
Automatic event location extraction from text plays a crucial role in many applications such as infectious disease surveillance and natural disaster monitoring. The fundamental limitation of previous work such as SpaceEval is the limited scope of extraction, targeting only at locations that are explicitly stated in a syntactic structure. This leads to missing a lot of implicit information inferable from context in a document, which amounts to nearly 40% of the entire location information. To overcome this limitation for the first time, we present a system that infers the implicit event locations from a given document. Our system exploits distributional semantics, based on the hypothesis that if two events are described by similar expressions, it is likely that they occur in the same location. For example, if 쏛 bomb exploded causing 30 victims and 쐌any people died from terrorist attack in Boston are reported in the same document, it is highly likely that the bomb exploded in Boston. Our system shows good performance of a 0.58 F1-score, where state-of-the-art classifiers for intra-sentential spatiotemporal relations achieve around 0.60 F1-scores.