Philippine Standard Time

Friday 24th of May 2024

Performance of genetic algorithm optimized Doc2Vec-kNN for classifying space science and adjacent fields documents with heterogenous sampling

Research and development (R&D) for space science and technology applications (SSTA), specially for those directed towards outer space science, require large amount of resources, and high level of expertise. While the Philippine space program is relatively young, the Philippines have a number of excellent scientists and engineers who work on fields that are related to SSTA, referred to as space adjacent fields. Support for space adjacent R&D will push the Philippines further in the field of SSTA, and help us in our goal of creating value from space for Filipinos and the world. The team from PhilSA developed a machine learning tool using a method called k-nearest neighbor (kNN) that will classify research works produced in the Philippines into twelve space adjacent areas important to outer space SSTA. Results show that the tool developed in PhilSA is able to identify the R&D field that a paper belongs to within acceptable accuracy. This tool will help PhilSA as it plans its future activities together with Philippine scientists and engineers.

PB-04 Hilario et al

This paper is part of the 41st Samahang Pisika ng Pilipinas Physics Conference (SPP 2023). These conference proceedings can be accessed for free via the SPP proceedings website upon account registration.