|
Article information
2026 , Volume 31, ¹ 4, p.83-92
Klimonov M.S., Fedorov R.K.
Sequential recommendation of web services for spatial data processing based on DAGNN
The paper addresses the problem of sequential recommendation of web services for spatial data processing on a geoportal. An approach based on graph neural network is proposed that takes into account both the structure of relations between services and the context of the current processing workflow. Service invocation logs and dependency graph are both used for training. Comparison with a frequency based baseline shows that the proposed model predicts the further service more accurately.
Keywords: recommender system, web service, geoportal, spatial data, graph neural network, DAGNN
Author(s):
Klimonov Mikhail Sergeevich
Position: Student
Office: Matrosov Institute for System Dynamics and Control Theory of Siberian Branch of Russian Academy of Sciences
Address: 664033, Russia, Irkutsk, Lermontov Str., 134
E-mail: mihail.klimonov@gmail.com
SPIN-code: 5410-2442Fedorov Roman Konstantinovich
Dr.
Position: Leading research officer
Office: Matrosov Institute for System Dynamics and Control Theory of Siberian Branch of Russian Academy of Sciences
Address: 664033, Russia, Irkutsk, Lermontov Str., 134
Phone Office: (3952) 45-31-08
E-mail: fedorov@icc.ru
SPIN-code: 5344-2226
Bibliography link:
Klimonov M.S., Fedorov R.K. Sequential recommendation of web services for spatial data processing based on DAGNN // Computational technologies. 2026. V. 31. ¹ 4. P. 83-92
|
|
|
|
