A semantic similarity computation method for virtual resources in cloud manufacturing environment based on information content. (April 2021)

• Record Type: Journal Article
• Authors: Zhang, Zhe; Chen, Youling; Wang, Xu
• Is Part Of: Journal of manufacturing systems. Volume 59(2021)
• Page Start: 646
• Page End: 660
• Publication Date: 2021-04
• Subjects: Cloud manufacturing -- Video semantic description -- Semantic similarity -- Information content -- Event cloud; Manufacturing processes -- Periodicals; Production engineering -- Data processing -- Periodicals; Robots, Industrial -- Periodicals; Production, Technique de la -- Informatique -- Périodiques; Robots industriels -- Périodiques; Electronic journals; 670.42
• DOI: 10.1016/j.jmsy.2021.04.014
• Languages: English
• ISSNs: 0278-6125

Abstract

Highlights: A semantic description frame for virtual resources based on video semantic description (VSD) is proposed. Event cloud (EC) is proposed to construct a virtual resource classification set, which reduces search scope drastically. A semantic similarity calculation model is implemented, which achieves a quantitative similarity calculation. Abstract: To ensure the effective and reasonable classification of virtual resources and provide support for the search and matching of required resources that accomplish tasks in cloud manufacturing (CMfg) environments, the semantic similarity computation of virtual resources needs to be promoted for that it is difficult to ensure that a large number of effective features participate in the calculation of semantic similarity and it is difficult to practically apply to efficient classification. However, to promote semantic similarity calculations still have problems such as difficulty in unifying the virtual resource pool and lagging in updating semantic depth calculations. In this paper, we propose a creative approach that can compute the semantic similarity of virtual resources described by video semantic description (VSD) based on information content (IC). In this approach, first, a multi-level progression semantics description framework for virtual resources is proposed with the help of VSD that can provide a new solution because of its explicit description and dynamic update to obtain real and reliable data; second, an event cloud (EC) that serves as the computing scope for semantics similarity calculation of virtual resources is constructed to settle the problem of substantial data growth caused by real-time resource updates on top of the resource manufacturing processes; third, based on an effective solution of corpus dependency and data sparseness, an extended semantic computing model is developed to calculate quantificationally the semantic similarities of virtual resources based on semantic relationships. The results show that compared with classifying virtual resources qualitatively, the semantic similarity computing method that starts from the manufacturing result and considers the comprehensive relationship is more reliable and effective.