Crowdsourcing website 集思廣益 is now officially on-line! We sincerely invite you to participate in putting human minds together! .
Dr. Wen-Lian Hsu and Academia Sinica team led by academician Dr. Chi-Huey Wong, who published their results AI empowered one-pot glyco-synthesis in Nature Communications. (2018).
Reference: Academia Sinica News
Congrats to the IASL research team and the IISR research team won the first prize in BioCreative V.5 Gene/Protein Related Object Recognition (GPRO) task.The IASL research team led by Dr. Wen-Lian Hsu, Director and Distinguished Research Fellow at the Institute of Information Science, collaborated with the IISR research team led by Dr. Richard Tzong-Han Tsai.
They won the first prize in BioCreative V.5 Gene/Protein Related Object Recognition (GPRO) task. The participants included Harbin Institute of Technology Shenzhen Graduate School, Dalian University of Technology, Goethe University Frankfurt, and other top research teams around the world.
Gene/protein plays a crucial role in biomedical research. Traditional gene/protein recognition tasks identify all genes and proteins related words/phrases. However, a gene/protein that can be linked to official database, such as EntrezGene or UniProt, is more useful for biomedical researchers. The goal of GPRO task is to detect them in patent titles and abstracts. In this competition, this research team used a novel approach called Statistical Principle-based Approach (SPBA), a pattern-based machine learning approach, to recognize gene- and protein-related objects with a higher recall. Besides, a Conditional Random Fields model is used to adjust the boundaries. Such an integrated framework led to their success in winning the first prize.
Team members: Po-Ting Lai, Ming-Siang Huang, Chu-Hsien Su, Dr. Richard Tzong-Han Tsai and Dr. Wen-Lian Hsu
Reference: BioCreative V.5 Challenge Evaluation Workshop
Congrats to Dr. Shih-Hung Liu for completing his PhD in electronic engineering from NTU!
Congrats to Dr. Yung-Chun Chang for completing his PhD in information management from NTU!
Congratulations to Dr. Wen-Lian Hsu for receiving the IICM Medal of Honor 2012 from Institute of Information & Computing Machinery.
Institute of Information & Computing Machinery(IICM) are a nonprofit organization devoting ourselves to fulfill the following purposes:
1. Promote IT related research and application in fields of engineering, science, management, education, and other disciplines.
2. Encourage academic and technical exchange between local groups and related international organizations.
3. Accelerate the advancement of IT in local communities.
Reference: IIS News
Congratulations to Cheng-Wei Shih, Chad Liu,Cheng-Wei Lee, and Wen-Lian Hsu for receiving the First place in NTCIR-10 Recognizing Inference in Text Task.Institute of Information & Computing Machinery(IICM) are a nonprofit organization devoting ourselves to fulfill the following purposes:
The Intelligent Agent Systems Lab (IASL), lead by Dr. Wen-Lian Hsu, won the first place in NTCIR-6Cross Language Question Answering (CLQA) System competition. It is the second year for such a competition in NTCIR. The IASL team, with years of experience in developing QA system, won the first place with a precision of 54.7% in the Chinese-Chinese mono-lingual subtask and 33.3% in the English-Chinese cross-lingual subtask, which is substantially better than last year's best results, 44.5% and 12%, respectively. Compared to QA systems in some other languages,the precision of Chinese QA system is lower. One of the main reasons is the unknown word and word segmentation problem in Chinese. IASL will continue the work trying to incorporate rules or templates and other semantic resources to improve the performance of Chinese QA systems. The NTCIR Workshop held by NII in Japan, is one of the most important international conferences of information processing in Asia. The first annual meeting of NTCIR workshop was held in 1999. The competitions of 2006 include Cross Language Information Retrieval (CLIR) Task, Patent Retrieval Task (PATENT), Question Answering Challenge (QAC), Multimodal Summarization (MuST) and Cross Language Question Answering (CLQA).
IASL team won the first place over 13 teams in the WS CityU closed track of the SIGHAN 2006 Word Segmentation ContestReference: SIGHAN bakeoff 2006
IASL team was ranked 2nd among 10 teams in the WS CKIP closed track of the SIGHAN 2006 Word Segmentation ContestReference: IASL team was ranked 2nd among 8 teams in the NER CityU closed track of the SIGHAN 2006 Named Entity Recognition Contest
Wen-Lian Hsu won the Academic Sinica Investigator AwardProf. Wen-Lian Hsu won the Academic Sinica Investigator Award in recognition of outstanding research by providing long-term full research support.
Prof. Wen-Lian Hsu has been nominated as an IEEE Fellow!The Fellow Committee has named 271 IEEE Senior Members to Fellow Grade effective 1 January 2006.
The new Fellows are listed below.
Prof. Wen-Lian Hsu
for contributions to natural language systems and bioinformatics.
Reference: Class of 2006 IEEE Fellows
IASL team was ranked 6th among 32 teams in the TREC Genomics Track Ad-hoc Retrieval Contest中研院資訊所許聞廉老師主持的「智慧型代理人系統實驗室」參加2005年基因體文件檢索比賽（TREC Genomics Track），Ad-hoc Retrieval競賽中，在與會32隊中排名第6。TREC為美國國家標準局（NIST）所主辦的國際會議，與會者多為資訊檢索相關學者，會議目的在促進相關領域技術交流。第一屆會議開始於1992，每年都舉辦多項資訊檢索比賽，今年的比賽項目包括企業文件檢索比賽（Enterprise Track）、基因體文件檢索比賽（Genomics Track）、高精準度檢索比賽（HARD Track）、問答比賽（Question Ansering Track）、Robust Retrieval Track、垃圾郵件過濾比賽（SPAM Track）等七項競賽。其中基因體文件檢索比賽已經是第三屆，而Ad-hoc Retrieval競賽則是第二屆。今年，基因體文件檢索比賽的題目混合了自然語言檢索題及表格式查詢，由於基因體文件查詢中牽涉到許多蛋白質及基因等數量大、無命名規律性的生物醫學名稱，因此要尋找到相關文件的困難度極高。智慧型代理人實驗室累積多年資訊擷取、資訊檢索、自然語言處理、生物資訊及問答系統開發經驗，在今年第一次參賽即獲得24.53%正確率，與美加兩地頂尖的York University、IBM、University of Waterloo、UIUC及National Library of Medicine（NLM）非常接近。該實驗室表示此次由於初次參加，僅做了關鍵字及關鍵詞片語擴展，未來將會加入更多生醫領域知識的運用，持續改進系統，提升檢索結果的正確率。
IASL team won the first place in the NTCIR5 CLQA Chinese Question Answering ContestThe Intelligent Agent System Lab (IASL), lead by Dr. Hsu, won the first place in NTCIR Cross Language Question Answering (CLQA) System competition. The NTCIR Workshop held by NII in Japan, is one of the most important international conferences of information processing in Asia. The first annual meeting of NTCIR workshop was held in 1999. The competitions of 2005 include Cross Language Information Retrieval (CLIR) Task, Patent Retrieval Task (PATENT), WEB Task (WEB), Question Answering Challenge (QAC), and Cross Language Question Answering (CLQA). It is the third year for QAC which is a competition of QA systems in Japanese. The QAC champion of last year is NTT, reaching the precision of 51.3%. This year, it is the first time for the NTCIR to hold CLQA. As to this reason, the questions this year, focus on people, location and artifacts, is simple and clear compared to other contests. IASL, with years QA system developing experience, won the first place with the precision of 44.5%. Compared to QA systems in some other languages, the precision of Chinese QA system remains relatively low according to this year's results of CLQA. One of the main reasons is the unknown word and word segmentation problem in Chinese. IASL will continue the work trying to incorporate rules or templates and other semantic resources to improve the performance of Chinese QA systems.
> IASL New Logo Announcement
IASL Logo (70X70)