SBIR-STTR Award

Traditional thesauri typically have entries with semantic relationships that are helpful for manual use, but the format of the semantic information is usually not conducive to their incorporation in Web retrieval applications. In addition, the number of thesaurus entries and relationships are limited, compared to the much larger and richer concept space of the scientific domain that a given thesaurus represents. With scientific and technological advances over time, it is desirable to develop automated means of capturing important new concepts and relationships, in order to make thesauri more adaptive and useful for searching. This project will develop technology for converting reference compilations, such as DOE's ETDE/INIS Joint Thesaurus, into a format useful for World Wide Web and Semantic Web search applications. The new Energy Semantic Network (ESN) will be used in conjunction with an advanced meta-search and clustering engine to automatically identify and categorize strongly associated concepts from the energy literature for potential inclusion in the thesaurus.

Commercial Applications and Other Benefits as described by the awardee:
The Energy Semantic Network not only should improve access to DOE/OSTI databases but also should be applicable to other scientific knowledge sources and retrieval applications

The lack of suitable data structures and ranking mechanisms hinder the effective utilization of semantic repositories in operational scientific and technical retrieval systems. Such systems could greatly benefit from semantic knowledge to improve the recall and precision of search queries. This project will create rich semantic networks that can be interfaced with search engines to enhance query formulation and increase the precision and recall of better organized search results. In Phase I, a prototype Energy Semantic Network was developed and integrated with a state-of-the-art meta-search and discovery engine that organizes the search results into meaningful semantic clusters to facilitate exploration and new discoveries. Phase II will create a capability to adaptively augment the Energy Semantic Network with new concepts and relationships. Then, the Energy Semantic Network will be tested, evaluated, and interfaced with DOE retrieval systems.

Commercial Applications and Other Benefits as described by the awardee:
The new semantic network technology should improve the quality and usefulness of search results from scientific and technical databases. In addition to the DOE application, the semantic network technology could be adapted to other scientific domains and technology applications