(76e) Forty Years of Computers and Chemical Engineering (1977-2017): Analysis of the Field Via Natural Language Processing Techniques

Recent advances in machine learning and natural language processing have produced a variety of tools for computerized mining of textual databases. However, while the mining of numerical data has enjoyed considerable attention, relatively little work has been done by chemical engineers in the analysis of textual information. In this work, we analyzed a large body of literature in chemical engineering using state-of-the-art text mining techniques. Particularly, the focus of attention is in process systems engineering and computer applications in the field of chemical engineering. Given its prominence and relevance to the field, the forty year old journal of Computers and Chemical Engineering (CACE) was investigated for our purposes.

In this work we are interested in answering following two questions. What are the main topics that have appealed to the CACE community since the journal was launched? How has interest in these topics evolved over time? These questions were addressed through topic modeling techniques, including non-negative matrix factorization and the structural topic model [1-4]. The results show that CACE covers diverse topics from process synthesis to artificial neural networks. A total of 18 research topics were identified when we maximized topic coherence [5]. Two dominant research topics that draw a great attention to the community are mathematical programming and process control. The topic prevalence analysis shows that the use of machine learning techniques can be traced back to 1990s. Additionally, we discovered that there is a close relationship between topics “artificial neural network” and “fault diagnosis.” These and other findings from our analysis will be discussed in detail.

References

[1] Greene, Derek, and James P. Cross. “Exploring the political agenda of the European parliament using a dynamic topic modeling approach.” Political Analysis 25.1 (2017): 77-94.

[2] Roberts, Margaret E., et al. “Structural Topic Models for Open-Ended Survey Responses.” American Journal of Political Science 58.4 (2014): 1064-1082.

[3] Roberts, Margaret E., Brandon M. Stewart, and Edoardo M. Airoldi. “A model of text for experimentation in the social sciences.” Journal of the American Statistical Association 111.515 (2016): 988-1003.

[4] Wang, Chong, and David M. Blei. “Variational inference in nonconjugate models.” Journal of Machine Learning Research 14.Apr (2013): 1005-1031.

[5] O’Callaghan, Derek, et al. “An analysis of the coherence of descriptors in topic modeling.” Expert Systems with Applications 42.13 (2015): 5645-5657.