Launch of the LODQA open source project

[Kevin B. Cohen]

Dear list members,

The Database Center for Life Science and University of Colorado School of Medicine are happy to announce the launch of the Linked Open Data Question-Answering (LODQA) project as an open-source project. We invite interested members of the community to interact with the current prototype system via our web-based interface, download the software, and consider joining the.

LODQA aims to explore interactions between Natural Language Processing and the Semantic Web by providing a platform for research and development regarding natural language question-answering from linked open data. The Semantic Web is rapidly approaching utility in many areas, including the life sciences. The dominant architecture for interacting with linked open data on the Semantic Web is SPARQL queries. However, SPARQL queries can be quite difficult to construct, even for experts. As a use case for NLP and the Semantic Web, LODQA provides a mechanism for generating SPARQL queries, given a question posed in natural language (currently English) as input. Thus, LODQA serves as a platform for research in question-answering as well as the Semantic Web, and additionally for intrinsic evaluation of a number of NLP enabling technologies, such as parsing, part of speech tagging, named entity recognition/normalization, and relation extraction.

The output of the project will be made available with an MIT open source license at

https://github.com/lodqa/lodqa

Please feel free to experiment with the current prototype system via a web-based interface at

http://lodqa.org/

Comments and participation are welcomed.

Jin-Dong Kim
Database Center for Life Science (DBCLS)
Research Organization of Information and  Systems (ROIS)
Tokyo, Japan

Kevin Bretonnel Cohen
Biomedical Text Mining Group
Computational Bioscience Program
University of Colorado School of Medicine
Aurora, Colorado