Identifying research trends of machine learning in business: a topic modeling approach

Abstract

Purpose This paper aims to discuss the suitability of topic modeling as a review method, identifies and compares the machine learning (ML) research trends in five primary business organization verticals. Design/methodology/approach This study presents a review framework of published research about adopting ML techniques in a business organization context. It identifies research trends and issues using topic modeling through the Latent Dirichlet allocation technique in conjunction with other text analysis techniques in five primary business verticals – human resources (HR), marketing, operations, strategy and finance. Findings The results identify that the ML adoption is maximum in the marketing domain and minimum in the HR domain. The operations domain witnesses the application of ML to the maximum number of distinct research areas. The results also help to identify the potential areas of ML applications in future. Originality/value This paper contributes to the existing literature by finding trends of ML applications in the business domain through the review of published research. Although there is a growth of research publications in ML in the business domain, literature review papers are scarce. Therefore, the endeavor of this study is to do a thorough review of the current status of ML applications in business by analyzing research articles published in the past ten years in various journals.

Exploring Technology- and Sensor-Driven Trends in Education: A Natural-Language-Processing-Enhanced Bibliometrics Study

“…The study identified 16 different topics, highlighting key themes such as "MOOC", "learning assessment", and "e-learning systems", which were consistently prevalent in the research literature. We found several additional studies addressing topic trend analyses in the scientific literature [29,30]. While these studies are usually focused on analyzing abstracts or only using text-based approaches, our research goes beyond the existing research by combining NLP techniques with a social network analysis, resulting in a more comprehensive and in-depth exploration of topic trends.…”

SCICERO: A deep learning and NLP approach for generating scientific knowledge graphs in the computer science domain

“…Then, common English stop-words provided by the NLTK 19 library as well as punctuation signs (e.g., apostrophes and commas) are removed. All entities are lemmatized using the WordNet lemmatizer 20 . In order to discard generic entities, SCICERO filters out the ones with a Information Content (IC) score 21 lower than a empirically defined threshold th IC (10 in the prototype, as reported in Section 4).…”

“…In 2021, we tackled these issues by introducing the information extraction approach described in Dessì et al [18], which is able to combine information from different tools according to a domain ontology and produce large-scale KGs. This approach inspired several further works in the field [19,20,21,22,23,24] and was used to produce the Artificial Intelligence Knowledge Graph (AI-KG) [25], a knowledge base describing 820K research entities in the field of AI. However, this first attempt also suffered from several limitations, such as: i) the entity extraction modules did not take advantage of the expert knowledge acquired from the analysis of the resulting knowledge graphs, ii) a limited ability to merge together multiple versions of the same entity (e.g., data cleaning algorithm, data cleaning automation, and preset data cleaning strategy), iii) a shallow and manual method for mapping verbal predicates to semantic relations, and iv) a limited methodology for assessing the validity of a triple, based on a very simple multilayer perceptron classifier.…”

Applications of Machine Learning (ML) - The real situation of the Vietnam Fintech Market

“…Cognition X report, London's role as an AI supplier5 base is twice the size of Berlin and Paris combined. London is also extremely well poised to be an AI and ML leader in finance and insurance, with new supplier formation growing at an annual rate of 42 per cent (compared with the global rate of 24 per cent annually (Pramanik & Jana, 2022).…”