Decision Support System for research project duplication detection

Decision Support Systems are usually computer applications along with a human component that can sift through large amounts of data and pick between many choices. Among many challenges in front of research managers, checking of duplication of research efforts is a major challenge. Existing systems do provide simple keyword base search wherein one can find similar projects. These systems do not take care of the semantics or sense of the word into consideration. Also, these systems fail to provide results if synonyms of the keywords are used in other research projects. Moreover, if the keywords are matched and are not used in the same sense, it is very much likely that the projects may not be similar. The presented approach in this work uses WordNet Ontology for finding the semantics of the textual information available within the system. Two algorithms viz. Filtered Set and Preferential Coefficient are presented that also take care the structure of the research project proposal. The algorithms are implemented for the PIMS-ICAR database that contains more than 5000 research project documents.