Abstract: [Purpose/Significance] The objective of this study is to develop a machine algorithm-based evaluation system for assessing the scholarly contributions of representative works, so as to further improve the accuracy and objectivity of peer review. [Design/Methodology] This study proposed an index system for assessing opinion contribution of the representative works. Then, by constructing a weighted exponential decay model and integrating the methods of view comparison and temporal order analysis, it comprehensively calculated these indexes. Additionally, a manual peer review was conducted, on the same sample dataset, and the consistency between the evaluation results obtained from manual assessment and the proposed model were compared, thereby validating the efficacy and practicability of the proposed method. [Findings/Conclusion] This study proposes a evaluation index model of opinion contribution combining thematic terms extraction and the comparison of Cosine Similarity. It is verified that the index value obtained from this model are highly consistent with the results of peer-review, reaching almost 86.85%, indicating that the model has high application value in assisting peer review. [Originality/Value] The TextRank4ZH model and time series methodology are integrated in this study to build an evaluation index model of opinion contributions that can be realized by the machine algorithm. This evaluation methodology can be used to assist the peer review system, and has strong practical significance in improving the reliability and impartiality of the evaluation results.

Keywords: Academic representative works, Peer review, Contribution degree, Evaluation model