=Paper=
{{Paper
|id=Vol-3745/paper13
|storemode=property
|title=Research on the Identification of Breakthrough Technologies Driven by Science
|pdfUrl=https://ceur-ws.org/Vol-3745/paper13.pdf
|volume=Vol-3745
|authors=Dan Wang,Xiao Zhou,Pengwei Zhao,Juan Pang,Qiaoyang Ren
|dblpUrl=https://dblp.org/rec/conf/eeke/WangZZPR24
}}
==Research on the Identification of Breakthrough Technologies Driven by Science==
https://ceur-ws.org/Vol-3745/paper13.pdf
Research on the Identification of breakthrough technologies
driven by science
Dan Wang 1, Xiao Zhou1,∗ , Pengwei Zhao1 , Juan Pang 1 and Qiaoyang Ren1
1 Xidian University, Xifeng 266 710126 Xi’an, Shaanxi Province, China
Abstract
The identification of breakthrough technologies plays a crucial role in driving technological innovation
forward. The science-driven technology innovation pattern has emerged as a significant approach for
identifying breakthrough technologies. This paper presents a novel framework for identifying
breakthrough technologies based on a science-driven technological breakthrough pattern. The
effectiveness of this framework is validated using the field of artificial intelligence as an illustrative
example. This method not only assists researchers in accurately identifying the sources and
development paths of technological breakthroughs but also provides important information for the
formulation of future research and development policies.
Keywords
Breakthrough technology, Knowledge networks, Link prediction, Structural entropy 1
1. Introduction enhancing national innovation capabilities and competi-
tiveness [5][6][7].
Breakthrough innovation, characterized by its highly This paper adopts a fine-grained representation ap-
revolutionary nature, plays a pivotal role in enabling en- proach, considering breakthrough technologies as com-
terprises to overhaul industry chains, enhance competi- posed of several closely related scientific and technolog-
tiveness, and seize prime opportunities in the increas- ical knowledge elements. To do so, this paper constructs
ingly competitive global landscape [1]. Recent research a breakthrough technology identification framework
has highlighted the significance of the interplay between based on the science-driven technology innovation pat-
science (S) and technology (T) in fostering potential tern. The core idea of the study is to use new science as
breakthrough technologies [2]. Scholars have started to a signal of innovation, to deeply explore the mechanisms
explore the complex correlation between S and T by in- and evolutionary paths through which new science leads
tegrating scientific literature and patent information. to technological breakthroughs, and on this basis, to
This integration has led to the identification of three pri- identify breakthrough technologies.
mary interaction patterns: science-driven (S-T), technol-
ogy-pull (T-S), and science-technology synergy (S&T).
Notably, the science-driven technology pattern signifies
2. Data and Method
instances where technological advancements stem from The framework for identifying breakthrough tech-
scientific discoveries, serving as a key driver of techno- nology is shown in Figure 1. Firstly, we use papers and
logical innovation [3][4]. The incorporation of scientific patents as carriers of science and technology, respec-
insights into technological progress plays a pivotal role in tively. We collect data from the Web of Science (WOS)
Joint Workshop of the 5th Extraction and Evaluation of Knowledge
Entities from Scientific Documents and the 4th AI + Informetrics (EEKE-
AII2024), April 23~24, 2024, Changchun, China and Online
∗
Corresponding author.
EMAIL: belinda1214@126.com (Xiao Zhou)
© Copyright 2024 for this paper by its authors. Use permitted under
Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
103
and Incopat patent databases, using search queries re- is used to partition the S-T revised network into 13 sub-
lated to the research topics to download relevant scien- networks. Two subnetworks that do not contain new sci-
tific papers and patents. Secondly, we focus on the ac- ence topics are excluded, leaving 11 subnetworks for fur-
quisition of new science, which is defined as scientific ther investigation.
topics that are both novel and impactful, yet have not We employ the structural entropy measure pro-
been integrated into existing technological systems. We posed by Xu et al. [13] to calculate the structural entropy
adopted Sentence-BERT (SBERT) [8] and Local Outlier influence of each subnetwork. We utilized the median as
Factor (LOF) [9] to quantify the novelty of papers, while a threshold and identified five subnetworks above this
utilizing citation counts as a metric for assessing paper median as potential breakthrough technologies. The fi-
impact. nal results were determined in conjunction with expert
Subsequently, we integrate new science into the ex- opinions. Ultimately, the study identified five break-
isting technological system through the construction of through technologies. Among them, drug discovery
a science-technology network. This network acts as a stands out due to its particularly significant impact. We
channel for merging new scientific findings with estab- conducted a detailed analysis of this breakthrough tech-
lished technological advancements. Link prediction is nology. Deep learning can train models using large-scale
employed to uncover deep semantic links between new biological data to predict the activity, toxicity, and other
science and technology. This is followed by the applica- properties of compounds, thereby rapidly screening can-
tion of community detection algorithms to filter subnet- didate drugs with potential therapeutic effects [14]. AI-
works containing new science-technology links. These discovered molecules were listed among the Massachu-
subnetworks serve as focal points for further analysis setts Institute of Technology (MIT)'s top ten break-
and evaluation. Finally, the impact of these subnetworks through technologies in 2020. In recent years, drug dis-
is evaluated using structural entropy to identify break- covery based on deep learning algorithms has gradually
through technologies. transitioned from research and development to practical
technology development. The from-scratch drug design
based on deep learning algorithms was recognized by
MIT as a breakthrough in successfully applying artificial
intelligence to the drug design process [15].
4. Discussion and Conclusion
This paper proposes a framework for identifying
breakthrough technology, starting with new sciences as
an innovation signal and tracking the evolution of tech-
Figure 1: Research framework for identifying break- nological breakthroughs stemming from them. The pri-
through technology
mary contributions of this study can be listed as follows.
First, this study proposes a novel method for identifying
3. Empirical analysis breakthrough technologies based on the innovation pat-
To assess the efficacy of the suggested approach, the tern of science-driven technological breakthroughs. This
domain of artificial intelligence (AI) is selected as a rep- approach enables dynamic tracking and measurement of
resentative case study. Following a methodology similar the innovation process triggered by new science. Second,
to that outlined by Tsay et al. [10] and subsequent re- it provides an in-depth characterization of the essence
moval of duplicate records, a total of 236,333 publica- and core features of new science. Furthermore, by em-
tions and 29,468 patents related to AI, published be- ploying a topic-based fine-grained approach, the study
tween 2014 and 2018, were identified. identifies breakthrough technologies, while also tracking
The science-technology network consists of 1,161 the dynamic interaction trajectories between new sci-
nodes and 62975 connecting edges, yielding a network ence and technology at the semantic level.
density of 0. 0935. We adopt an attribute feature-based Several limitations of our proposed method require
graph convolutional network (GCN) [11] for link predic- further improvement. This paper primarily considers the
tion in the science-technology network to discover po- driving effect of science on technological breakthroughs.
tential linkages between new science topics and techno- Future research could explore the identification of
logical topics. After link prediction, Liu et al.'s method [12] breakthrough technologies under different patterns of
science and technology interaction. Moreover, alongside
104
scientific influence, the commercial aspect warrants at- [12] Y. Liu, M. Ott, N. Goyal, J. Du, M. Joshi, D. Chen, O.
tention. Levy, M. Lewis, L. Zettlemoyer, V. Stoyanov,
Roberta: A robustly optimized bert pretraining
5. Acknowledgements approach. ArXiv preprint arXiv, 1907: 11692, 2019.
[13] H. Xu, et al, A methodology for identifying
This work was supported by the General Program of breakthrough topics using structural
National Natural Science Foundation of China (Grant No. entropy. Information Processing & Management,
72374165) . 59(2): 102862, 2022.
[14] B. Bhinder, C. Gilvary, N. S. Madhukar, O. Elemento,
6. References Artificial intelligence in cancer research and
precision medicine. Cancer discovery, 11(4): 900-
[1] D. S. Hain, J. L. Christensen, Capital market
915, 2021.
penalties to radical and incremental innovation.
[15] Singh S, Gupta H, Sharma P, Sahi S, Advances in
European Journal of Innovation Management, Artificial Intelligence (AI)-assisted approaches in
23(2): 291-313, 2019. drug screening. Artificial Intelligence Chemistry,
[2] Z. Ba, K. Meng, Y. Ma, Y. Xia, Discovering 2(1): 100039, 2024.
technological opportunities by identifying dynamic
structure-coupling patterns and lead-lag distance
between science and technology. Technological
Forecasting and Social Change, 200: 123147, 2024.
[3] A. D. Malva, S. Kelchtermans, B. Leten, R.
Veugelers, Basic science as a prescription for
breakthrough inventions in the pharmaceutical
industry. The Journal of Technology Transfer, 40:
670-695, 2015.
[4] R. Veugelers, J. Wang, Scientific novelty and
technological impact. Research Policy, 48(6): 1362-
1372, 2019.
[5] J. Dedrick, K. L. Kraemer, Who captures value from
science-based innovation? The distribution of
benefits from GMR in the hard disk drive industry.
Research Policy, 44(8): 1615-1628, 2015.
[6] V. Narayanamurti, J. Y. Tsao, How technoscientific
knowledge advances: A Bell-Labs-inspired
architecture. Research Policy, 53(4): 104983, 2024.
[7] Q. Ke, An analysis of the evolution of science-
technology linkage in biomedicine. Journal of
Informetrics, 14(4): 101074, 2020.
[8] N. Reimers, I. Gurevych, Sentence-bert: Sentence
embeddings using siamese bert-networks. ArXiv
preprint arXiv, 1908: 10084, 2019.
[9] D. Jeon, J. Lee, J. M. Ahn, C. Lee, Measuring the
novelty of scientific publications: A fastText and
local outlier factor approach. Journal of
Informetrics, 17(4): 101450, 2023.
[10] M. Y. Tsay, Z. W. Liu, Analysis of the patent
cooperation network in global artificial intelligence
technologies based on the assignees. World Patent
Information, 63: 102000, 2020.
[11] T. N. Kipf, M. Welling, Semi-supervised
classification with graph convolutional networks.
In: Proceedings of ICLR 2017 (pp. 1–14).2017.
105