Increase in the amount of publicly available online data provides a rich resource for information extraction. Especially news web sites and social media posts are most available candidates of such resources. In this project, we aim to work on representations of such online resources in graph structure and to perform information extraction on these graphs. Graph data structure provides means to represent rich and complex relationships between different entities and this makes it possible to extract new and different kind of information.
The main objectives of the project are as follows:
The project team includes researchers from three different universities (METU, Gebze Technical University, Kocaeli University). Each researcher will contribute to the project and the graph database library in a complementary way on the basis of their expertise. Hence we consider the library to be developed to consists of three sub groups of capabilities such that each university will be responsible from one group. Additionally, project team includes a partner from industry, Somera. Somera will be in charge of testing the library by using it within their own problems and processing, and they will provide guidance and suggestions for improvement.
This work is supported by TUBITAK Ardeb 1003 program with 117E566 project number.
METU Computer Engineering Department
Kocaeli University Computer Engineering Department
Gebze Technical University Instıtude of Information Technologies Department