Abstract: 【Purpose/significance】This paper studies the measurement method of novelty of academic conference papers and provides
a new perspective for the novelty evaluation of conference papers.【Method/process】This paper designs a method to measure absorption
novelty index and output novelty index, and selects conference papers in the field of "artificial intelligence" for empirical research.
【Result/conclusion】Papers with high absorption novelty and high output novelty have the highest probability of becoming A-type con⁃
ference papers. The results show the effectiveness of the novelty measurement method, which has certain reference significance for the
automatic review of conference papers.【Innovation/ limitation】This paper designs the novelty measurement method of conference pa⁃
pers to promote the development of the evaluation of conference papers; But there are some domain limitations.