We built a semantic search engine for research grant applications
Until now, there was no repository of applications for grants submitted in competitions for grants for science and research projects, as well as of those projects which received funding. It was difficult for scientists submitting applications to find the information about previous projects in the same category or applications which would pertain to a similar topic. The answer to this problem is the semantic search engine recently built at our Institute.
Our search engine allows the search of project proposals in accordance with specified criteria in terms of applications’ text data, such as – for example – a summary or project description. It also has an intelligent component which uses the tools of artificial intelligence, natural language processing, and machine learning methods (grouping by subject/classification) allowing the search for proposals and projects with a similar thematic range, even if they were submitted for different competitions. Our new tool also enables a user to classify proposals in accordance with various criteria which can be a valuable support in the process of application evaluation.
Apart from the application search, the search engine allows one to access information pertaining to a specific application, it also has functionalities enabling the filtering and narrowing down search results, it also offers statistics regarding the applications meeting the search criteria requirements. Search results are presented in three ways: as a result list, thematic clusters (text groups), and infographics (tables, charts).
The aim behind building the semantic search engine was allowing society, potential applicants, and institutions that organize competitions managed by the electronic proposal submission system (OSF) functional access to information on competition applications, and allocated funding. I am pleased we have achieved it all – says Marek Kozłowski, Ph.D., the Manager of the Natural Language Processing Laboratory.
The semantic search engine was built as part of The Integrated System of Information on Science and Higher Education – Phase 1 project. The search engine can be accessed here: https://wyszukiwarkazsun.opi.org.pl.
Graphics: 123 rf