Role of Interdisciplinarity in Computer Sciences: Quantification, Impact and Life Trajectory

TL;DR

This study analyzes the evolution and impact of interdisciplinarity in computer science over 50 years, revealing how balanced interdisciplinary work maximizes scientific influence and how citation dynamics shape research field trajectories.

Contribution

It introduces a comprehensive quantification of interdisciplinarity, analyzes its effect on impact, and uncovers the life cycle phases of research fields in computer science.

Findings

Balanced interdisciplinarity correlates with highest impact.

Highly disciplinary or interdisciplinary papers tend to have lower impact.

Research fields follow a three-phase life trajectory: growth, maturity, and interdisciplinarity.

Abstract

The tremendous advances in computer science in the last few decades have provided the platform to address and solve complex problems using interdisciplinary research. In this paper, we investigate how the extent of interdisciplinarity in computer science domain (which is further divided into 24 research fields) has changed over the last 50 years. To this end, we collect a massive bibliographic dataset with rich metadata information. We start with quantifying interdisciplinarity of a field in terms of the diversity of topics and citations. We then analyze the effect of interdisciplinary research on the scientific impact of individual fields and observe that highly disciplinary and highly interdisciplinary papers in general have a low scientific impact; remarkably those that are able to strike a balance between the two extremes eventually land up having the highest impact. Further, we study the reciprocity among fields through citation interactions and notice that links from one field to related and citation-intensive fields (fields producing large number of citations) are reciprocated heavily. A systematic analysis of the citation interactions reveals the life trajectory of a research field, which generally undergoes three phases -- a growing phase, a matured phase and an interdisciplinary phase. The combination of metrics and empirical observations presented here provides general benchmarks for future studies of interdisciplinary research activities in other domains of science.