2024-03-29T08:25:42Zhttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_oaipmhoai:ipsj.ixsq.nii.ac.jp:002072242023-04-27T10:00:04Z01164:08228:10133:10353
Towards Automated Generation of Data Models A Case Study of ECHONET Device Objects SpecificationTowards Automated Generation of Data Models A Case Study of ECHONET Device Objects Specificationenghttp://id.nii.ac.jp/1001/00207122/Technical Reporthttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_action_common_download&item_id=207224&item_no=1&attribute_id=1&file_no=1Copyright (c) 2020 by the Information Processing Society of JapanJapan Advanced Institute of Science and TechnologyJapan Advanced Institute of Science and TechnologyUniversity of Science Ho Chi MinhJapan Advanced Institute of Science and TechnologyVan, Cu PhamThanh, Tung LeTien, Huy NguyenYasuo, TanIn accordance with the development of cloud computing, the term Web APIs is becoming the backbone of the Internet of Things. The publication of data models is extremely important to enhance the interoperability of Web APIs providers and Web APIs consumers. Currently, data models of a protocol are created by experts, however, humans might cause mistakes such as syntax errors, typographical errors, and inconsistencies of abbreviations. This paper introduces an idea of the automated generation of data models by software (machine) as it can handle the drawbacks of humans and it is much faster than humans. The biggest barrier of using the machine is that it lacks domain knowledge to export terms and abbreviations from protocol descriptions in natural languages. To this end, Natural Language Processing models are utilized. In this research, a case study of the ECHONET Lite protocol is introduced. The results proved that machines can mimic experts in exporting terms and abbreviations while assures the syntactic error-free and consistency of generated data models. Furthermore, the machine supports fast, reliable while enhances reusability in exporting data models for multiple platforms.In accordance with the development of cloud computing, the term Web APIs is becoming the backbone of the Internet of Things. The publication of data models is extremely important to enhance the interoperability of Web APIs providers and Web APIs consumers. Currently, data models of a protocol are created by experts, however, humans might cause mistakes such as syntax errors, typographical errors, and inconsistencies of abbreviations. This paper introduces an idea of the automated generation of data models by software(machine) as it can handle the drawbacks of humans and it is much faster than humans. The biggest barrier of using the machine is that it lacks domain knowledge to export terms and abbreviations from protocol descriptions in natural languages. To this end, Natural Language Processing models are utilized. In this research, a case study of the ECHONET Lite protocol is introduced. The results proved that machines can mimic experts in exporting terms and abbreviations while assures the syntactic error-free and consistency of generated data models. Furthermore, the machine supports fast, reliable while enhances reusability in exporting data models for multiple platforms.AA1271737X研究報告高齢社会デザイン(ASD)2020-ASD-1911172020-09-222189-44502020-10-01