Measuring Technology Complementarity Between Enterprises With an hLDA Topic Model

Wang, Xuefeng; Qiao, Yali; Hou, Yujia; Zhang, Shuo; Han, Xiaotong

doi:10.1109/tem.2019.2958113

Cited by 7 publications

(1 citation statement)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…2) Topic Modeling: In contrast to search string refinement, which comes from the information retrieval domain, topic modeling belongs to the machine learning domain. In recent years, topic modeling has been widely used to analyze documents or texts for various purposes, e.g., by [26], for the identification of breakthrough innovations, by [27], to measure the technology complementarity of companies, or by [28], to discover technological opportunities. In the following, we outline some of its basic ideas.…”

Section: B Methodical Considerations For the Evaluation Stepmentioning

confidence: 99%

Stimulating R&D by Finding Frugal Patents: A Process Model and a Comparison Between Different Evaluation Methods

Kronemeyer

Draeger

Moehrle

2023

IEEE Trans. Eng. Manage.

View full text Add to dashboard Cite

In order to stimulate research and development in their companies, managers, and technical experts might aim at identifying frugal patents from third parties by means of advanced text mining techniques. The same may be true for scientists who wish to improve their understanding of frugal inventions. For this purpose, we present a process model in four steps, and particularly analyze the evaluation step by applying different text-mining methods, namely search string refinement and topic modeling. Testing the process model in the technological field of white goods, we find that the process model as a whole as well as both evaluation methods are able to detect frugal patent candidates. Nevertheless, both methods have different profiles in terms of precision and recall. As a theoretical contribution, we suggest differentiating between two types of frugal invention namely those with frugal attributes concentrated in parts of the description and those with frugal attributes distributed throughout the description. As a managerial contribution, we show that a hybrid solution may be effective in identifying frugal patent candidates while reducing manual efforts for evaluation at the same time.

show abstract

Section: B Methodical Considerations For the Evaluation Stepmentioning

confidence: 99%