Creating Knowledge Out of Interlinked Data<br />KAIST Project<br />Mun Y. Yi<br />19-09-2011<br />
Agenda<br />Introduction of KAIST<br />KAIST LOD Team<br />Description of Work<br />Tasks<br />Deliverables<br />Current S...
Introduction of KAIST<br />KAIST (Korea Advanced Institute of Science and Technology) is the first and top science and tec...
KAIST LOD Team<br />Key-Sun Choi<br />Director of Semantic Web Research Center<br />Head of the Computer Science Departmen...
Work Description: Tasks<br />Task 3.2: Provenance-Aware Linked Data Extraction from Unstructured and Semi-Structured Sourc...
Work Description: Deliverables<br />Deliverable 3.2.4 Korean NLP2RDF (KAIST, M32)<br />Initial release of the NLP2RDF fram...
Current Status<br />In preparation for a proposal to Korea MKE (Korea Ministry of Knowledge and Economy)<br />Need to invo...
Upcoming SlideShare
Loading in...5
×

LOD2 Plenary Meeting 2011: KAIST – Partner Introduction

730

Published on

Slides of KAIST (Korea) for their partner introduction as a new LOD2 partner in the course of the LOD2 project enlargement - presented at the LOD2 plenary meeting in Leuven, Belgium on September 2011

Published in: Business, Education
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
730
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
4
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Transcript of "LOD2 Plenary Meeting 2011: KAIST – Partner Introduction"

  1. 1. Creating Knowledge Out of Interlinked Data<br />KAIST Project<br />Mun Y. Yi<br />19-09-2011<br />
  2. 2. Agenda<br />Introduction of KAIST<br />KAIST LOD Team<br />Description of Work<br />Tasks<br />Deliverables<br />Current Status<br />
  3. 3. Introduction of KAIST<br />KAIST (Korea Advanced Institute of Science and Technology) is the first and top science and technology research university in Korea. <br />Founded in 1971 to raise elites in science and technology<br />Located in the Daedeok Research Complex in the city of Daejeon, 150 kilometers south of Seoul.<br />For the 2009 academic year, over 8000 students enrolled; 3452 in the bachelor’s, 2197 in the master’s, and 2357 in the doctorate program. KAIST has 842 professors and 334 staff members as of January 2009<br />According to QS World University Rankings 2011, KAIST is ranked as the 90th in the World and 2nd in Korea.<br />
  4. 4. KAIST LOD Team<br />Key-Sun Choi<br />Director of Semantic Web Research Center<br />Head of the Computer Science Department<br />Expertise in ontology, NLP, and semantic Web<br />Mun Y. Yi<br />Director of Knowledge Systems Lab<br />Associate professor in the Knowledge Service Engineering Department<br />Expertise in knowledge engineering, recommender systems, e-learning, and MIS/HCI<br />In-Young Ko<br />Director of WebEng Lab<br />Associate professor in the Computer Science Department<br />Expertise in software engineering and Web engineering including Web services, Web-based information management, and semantic Web<br />Ying Liu<br />Director of Intelligent System and Service Lab<br />Assistant professor in the Knowledge Service Engineering Department<br />Expertise in Tableseer, information retrieval, and text mining<br />
  5. 5. Work Description: Tasks<br />Task 3.2: Provenance-Aware Linked Data Extraction from Unstructured and Semi-Structured Sources <br />KAIST will add its experience in extracting Linked Data from Korean resources. KAIST has the most advanced technology in processing Korean natural language resources and data. One example of such resource is CoreNet, which contains a taxonomic hierarchy, concept definitions and frame sets for Korean, Japanese and Chinese words. KAIST will build a Korean version of NLP2RDF by integrating various Korean natural language tools and providing the result of those toolkits in RDF format. KAIST will also facilitate the standardization of NLP2RDF through its involvement in the ISO group TC37/SC4 (Language Resources Management). <br />Task 4.1: Semi-Automatic Data Interlinking<br />KAIST will contribute to this task by providing a platform for automatic linking with Korean, Chinese, Japanese RDF resources. CoreNet contains a hierarchical concept structure for Korean, Chinese and Japanese words. Once the concepts of CoreNet are mapped to WordNetsynsets, as WordNet is already integrated into LOD, KAIST can provide the Korean, Chinese and Japanese RDF data integration platform for Linked Data by providing a mapping mechanism of those data to CoreNet, thus solving multilingual issues for these Asian languages. KAIST has taken the initial step of the CoreNet-WordNet mapping; already showing some progress<br />Task 4.5a: Multilingual Linked Data Fusion <br />KAIST will choose the DBpedia dataset as the pivot multilingual dataset, since it is extracted from various kinds of languages. KAIST will work on the multilingual fusion of those multilingual DBpediadatasets, thus eliminating issues for other multilingual resources, since they simply need to fuse with their own language DBpedia resource. As a first step, KAIST is working on the bilingual fusion between the Korean DBpedia and the English DBpedia; having already obtained some results. At the end of the project these results will be expanded to the fusion of Chinese and Japanese DBpedia with Korean and English DBpedia. We envision to reach more than 90% precision and recall with this multi-lingual fusion approach. <br />Task 6.4: Development of application scenarios and testing of the LOD2 stack configurator<br />The stack configurator will enable potential users to create their own personalized version of the LOD2 Stack, which contains only those functions relevant for their usage scenarios. In this task, LOD2 partners will conduct an in-depth analysis of different application scenarios and identify LOD2 functional components that adequately respond to specific application requirements. These results of the study will be used to assist the development of the stack configurator and to prepare comprehensive LOD2 documentation both from the engineer’s and the user’s viewpoint.<br />Task 10.2d: Training and Dissemination in Korea (KAIST). <br />KAIST will ensure the penetration of LOD2 results in a dynamic Asian country by organizing a number of events and outreach activities, such as: <br />Two research-oriented Data Web symposia aiming to bring together relevant researchers in Asia with the LOD2 consortium, <br />Two industry workshops aiming at disseminating LOD2 results to Korean and Japanese companies and to facilitate cooperation and market entry of industrial LOD2 partners, <br />One Asian Data Web summer school aiming to outreach to PhD students and young researchers. <br />
  6. 6. Work Description: Deliverables<br />Deliverable 3.2.4 Korean NLP2RDF (KAIST, M32)<br />Initial release of the NLP2RDF framework for Korean text. This will include various Korean NLP tools and data, including CoreNet. Compared to English, Korean NLP toolkits are less developed and opened; hence, most of the time will be devoted to the new development of Korean NLP tools which will contribute to LOD. <br />Deliverable 4.1.3 Korean Resource Linking Assist Release (M24)<br />The first version of Korean resource linking assist to DBpedia will intelligently recommend and order the possible mappings to the knowledge engineer. This will be implemented as the expansion of Deliverable 4.1.1. <br />Deliverable 4.1.4 Asian Resource Linking Assist Release (M30)<br />This tool will help the knowledge engineer to link Korean, Chinese, Japanese language resources to Linked Data by recommending and ordering appropriate mappings to her. <br />Deliverable 4.5.3 Korean Data Fusion Assistant (M30)<br />The component will support Korean data fusion into English LOD by combining Deliverable 4.5.1 with the fused dataset of English and Korean DBpedia. More precisely, the component will first fuse the new Korean dataset into Korean DBpedia by using D4.5.1, and the result will again be fused into the English DBpedia by applying the fusion result of Korean and English DBpedia. <br />Deliverable 4.5.4 Asian Data Fusion Assistant (M36)<br />The component is an extension of Deliverable 4.5.3, and will support the data fusion of Korean, Japanese and Chinese datasets. <br />
  7. 7. Current Status<br />In preparation for a proposal to Korea MKE (Korea Ministry of Knowledge and Economy)<br />Need to involve industry partners<br />Potential projects/applications<br />CoreNet to LOD<br />Korean NLP2RDF<br />Multilingual DBPedia matching and expansion<br />Link Korea Traditional Knowledge DB to LOD<br />Have similar work done in China and Japan<br />Wiki History and Wiki Q&A<br />Korean Wiki annotation<br />
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×