There is a growing trend towards combining mathematics and computer science education. Although there are surely synergies between the two disciplines, has sufficient thought been given to the benefit of fostering independently the unique skillsets they offer in order to best harness these synergies? By comparing and contrasting quantitative evidence relating to skillsets demonstrated through choice of words in describing research contributions to a field common to both mathematics and computer science, algorithms, this research aims to provide a better understanding of what are the unique differences in skillsets offered by the two disciplines. Through the use of natural language processing and machine learning techniques, applied in quantitative experimentation, it was found that phrases used by mathematicians to describe their research contributions to the field of algorithms showed a clear emphasis on (developing algorithms as a) method to finding a solution by (developing) functions and/ or convergent numerical methods (and validating them through) numerical examples. While phrases used by computer scientists showed an emphasis on using algorithms as an) approach to solve a problem with data (to train a) model (typically via) optimisation methods or numerical experiments, (and typically being a) machine learning (algorithm) such as a deep neural network or other neural network, reinforcement learning or deep learning, and on (the algorithm’s) performance and time to solve. This research fills a gap in the literature by providing quantitative experiment-based evidence of some of the unique differences in skillsets of mathematicians and computer scientists (on a high level, the skillsets required to develop effective algorithms to solve problems versus the skillsets to use (and experiment with) algorithms using a computer on data to solve problems in an optimal way). It is expected that this research would be of interest to educators and hiring managers who require to cultivate and leverage these skillsets to the benefit of society.
This data was collected using the arxiv.py Python wrapper for the arXiv API  with query term: "algorithm OR algorithms". The data corresponds to submissions to arXiv between the end of May 2022 and the beginning of December 2022. Columns correspond to the following attributes of the search results: