Technology Opportunity Analysis: Combining SAO Networks and Link Prediction

Detecting the first signs of change in one's technological surroundings is a critical factor in the success of an enterprise, and technology opportunity analysis can be a crucial process in identifying those signs. However, the common keyword-based methods of analysis do not fully express the r...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:IEEE transactions on engineering management Ročník 68; číslo 5; s. 1288 - 1298
Hlavní autoři: Han, Xiaotong, Zhu, Donghua, Wang, Xuefeng, Li, Jia, Qiao, Yali
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York IEEE 01.10.2021
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:
ISSN:0018-9391, 1558-0040
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Detecting the first signs of change in one's technological surroundings is a critical factor in the success of an enterprise, and technology opportunity analysis can be a crucial process in identifying those signs. However, the common keyword-based methods of analysis do not fully express the relationships between technologies. Subject-action-object (SAO) analysis offers a solution to this problem but, currently, these methods only consider the relationships that already exist. Yet, intuitively, technology opportunities are most likely to reside in potential connections. To test this notion, in this article we conduct a case study on malignant melanoma of the skin. First, we construct an SAO network of the titles and abstracts of medical documents, then use a link prediction algorithm to identify probable future links between unconnected nodes. These possible new technology combinations are further analyzed with a backtracking algorithm to reveal the most promising technology opportunities. Further analysis of the results combined with medical knowledge confirms the effectiveness of our method.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0018-9391
1558-0040
DOI:10.1109/TEM.2019.2939175