Abstract:Author impact evaluation and prediction play a key role in determining rewards, funding, and promotion. In this paper, we first introduce the background of author impact evaluation and prediction. Then, we review recent developments of author impact evaluation, including data collection, data pre-processing, data analysis, feature selection, algorithm design, and algorithm evaluation. Thirdly, we provide an in-depth literature review on author impact predictive models and common evaluation metrics. Finally, we look into the representative research issues, including author impact inflation, unified evaluation standards, academic success gene, identification of the origins of hot streaks, and higher-order academic networks analysis. This paper should help the researchers obtain a broader understanding in author impact evaluation and prediction, and provides future research directions.
What problem does this paper attempt to address?
The problems that this paper attempts to solve mainly focus on the field of academic impact assessment and prediction. Specifically, the paper is concerned with how to more accurately assess and predict the academic influence of authors, which is of great significance in determining rewards, fund allocation, and promotion, etc. The paper points out that with the advent of the big data era, the amount of academic data has grown exponentially. Although this has brought rich research resources, it has also brought unprecedented challenges. For example, it has become difficult to identify the most relevant research work or scholars from a large amount of academic data, and decision - makers need more information to support the research evaluation system, which should not only reflect past performance but also predict potential research productivity.
To meet these challenges, the paper reviews the research progress in author influence assessment and prediction in recent years, including multiple aspects such as data collection, pre - processing, analysis, feature selection, algorithm design, and evaluation, and deeply discusses the evaluation methods based on network structure and the prediction methods based on models. In addition, the paper also explores several representative problems in current research, such as author influence inflation, unified evaluation standards, academic success genes, identification of the origin of the popular stage, and high - order academic network analysis, etc. These problems are crucial for more accurately discovering academic rising stars and reasonably assessing and predicting the influence of authors.
In summary, this paper aims to help researchers obtain a broader understanding of author influence assessment and prediction by comprehensively reviewing existing research results and provide guidance for future research directions.