A novel technology life cycle analysis method based on LSTM and CRF

Jianhua Hou (), Shiqi Tang () and Yang Zhang ()
Additional contact information
Jianhua Hou: Sun Yat-sen University
Shiqi Tang: Sun Yat-sen University
Yang Zhang: Sun Yat-sen University

Scientometrics, 2024, vol. 129, issue 3, No 1, 1173-1196

Abstract: Abstract Technology life cycle (TLC) analysis provides essential support for investment-related strategies and helps to technology trajectory tracing, forecasting, and assessment. The most typical method used to identify TLC is the S-curve fitting method. However, doubts about its accuracy and reliability have been raised owing to the single indicator problem and the missing link between TLC and indicators. K-nearest neighbors (KNN) and hidden Markov model (HMM)-based methods are two influential methods that have been developed. However, something could be improved with these methods. The emerging order of stages is not under control, and the impact of early technology development on the later stages has yet to be addressed. These issues led us to propose a new method to identify TLC using multiple indicators based on machine learning techniques. We extracted ten indicators from the incoPat patent database and utilized a long short-term memory (LSTM) network–conditional random field (CRF) to identify TLC stages with the probability of technology being in a particular stage at a point of the year and changing to other stages during the following year. Moreover, this study investigates the theoretical meaning and empirical performance of indicators. 3-Dimensional print technology was selected as a case study, and its TLC was analyzed and prospects discussed. Comparison of this method and other methods are made as well. The results of our method that fit with the actual progression of technology are relatively accurate. Our analysis showed that the proposed method could offer a smooth and stationary TLC pattern that is accurate and easily understood.

Keywords: Technology life cycle; Technology progression; Long short-term memory network; Conditional random field; Multiple patent indicators (search for similar items in EconPapers)
Date: 2024
References: Add references at CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s11192-024-04946-z Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:scient:v:129:y:2024:i:3:d:10.1007_s11192-024-04946-z

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/11192

DOI: 10.1007/s11192-024-04946-z

Access Statistics for this article

Scientometrics is currently edited by Wolfgang Glänzel

More articles in Scientometrics from Springer, Akadémiai Kiadó
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().