Web services classification via combining Doc2Vec and LINE model
by Hongfan Ye; Buqing Cao; Jinkun Geng; Yiping Wen
International Journal of Computational Science and Engineering (IJCSE), Vol. 23, No. 3, 2020

Abstract: Classifying web services with similar functionality from tremendous amount of web services can significantly improve the efficiency of service discovery. Few of the web services classification researches integrate the independent mining of the content semantic information and network structure information hidden in the web service characterisation documents. Therefore, we propose a web service classification method combining them. So, the Doc2Vec algorithm is firstly exploited to deeply mine the functional semantics of web service characterisation documents and obtain web service's content semantic representation. Then, the LINE algorithm is adopted to embed the web service information network which is constructed by utilising web service characterisation documents and word information. Subsequently, the content semantic representation and network structure representation of web service are integrated as the input of the logistic regression classifier to perform web service classification. The experimental results on the ProgrammableWeb dataset verify that the proposed method outperforms to baseline methods.

Online publication date: Thu, 26-Nov-2020

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Computational Science and Engineering (IJCSE):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com