Home // SEMAPRO 2015, The Ninth International Conference on Advances in Semantic Processing // View article

Application of the Tensor-Based Recommendation Engine to Semantic Service Matchmaking

Authors:
Andrzej Szwabe
Michal Ciesielczyk
Pawel Misiorek
Michal Blinkiewicz

Keywords: Semantic Service Selection; tensor-based multirelational data modeling

Abstract:
The paper presents a novel approach to semantic Web service matchmaking, which involves a use of multilinear data representation and processing. The proposed solution involves the use of a novel tensor data filtering method based on a set of covariance matrices derived from a hierarchical tensor structure. We provide results of experimental evaluation of the proposed solution conducted with the use of the Semantic Service Selection (S3) contest dataset. The evaluation has been done using the standard Information Retrieval methodology that assumes the methodologically correct partitioning of the dataset on mutually exclusive subsets: the training set and the testing set. The experimental evaluation results presented in the paper indicate superiority of the covariance-based tensor filtering method over other state-of-the-art tensor processing methods in terms of the matchmaking quality measured using mean average precision and Area Under the ROC curve (AUROC) measures.

Pages: 116 to 125

Copyright: Copyright (c) IARIA, 2015

Publication date: July 19, 2015

Published in: conference

ISSN: 2308-4510

ISBN: 978-1-61208-420-6

Location: Nice, France

Dates: from July 19, 2015 to July 24, 2015