Augmenting research cooperation in production engineering with data analytics

Abstract

Understanding how members of a research team cooperate and identifying possible synergies may be crucial for organizational success. Using data-driven approaches, recommender systems may be able to find promising collaborations from publication data. Yet, the outcome of scientific endeavors (i.e. publications) are only produced sparingly in comparison to other forms of data, such as online purchases. In order to facilitate this data in augmenting research cooperation, we suggest to combine data-driven approaches such as text-mining, topic modeling and machine learning with interactive system components in an interactive visual recommendation system. The system leads to an augmented perspective on research cooperation in a network: Interactive visualization analyzes, which cooperation could be intensified due to topical overlap. This allows to reap the benefit of both worlds. First, utilizing the computational power to analyze large bodies of text and, second, utilizing the creative capacity of users to identify suitable collaborations, where machine-learning algorithms may fall short.