A social voting approach for scientific domain vocabularies construction

Scientific domain vocabularies play an important role in academic communication and lean research management. Confronted with the dramatic increasing of new keywords, the continuous development of a domain vocabulary is important for the domain to keep its long survival in the scientific context. Current methods based either on statistical or linguistic approaches can automatically generate vocabularies that consist of popular keywords, but these approaches fail to capture high-quality standardized terms due to the lack of human intervention. Manual methods take use of human knowledge, but they are both time-consuming and expensive. In order to overcome these deficiencies, this research proposes a novel social voting approach to construct scientific domain vocabularies. It integrates automatic system and human knowledge based on the theory of linguistic arbitrariness and selects widely accepted standardized set of keywords based on social voting. A social voting system has been implemented to aid scientific domain vocabulary construction in the National Natural Science Foundation of China. Two experiments are conducted to demonstrate the effectiveness and validity of the built system. The results show that the constructed domain vocabulary using this system covers a wide range of areas under a discipline and it facilitates the standardization of scientific terminology.