期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2019
卷号：97
期号：2
页码：633-643
出版社：Journal of Theoretical and Applied
摘要：Internet for e-commerce is the main source of information, this information is not directly exploitable by computers, hence many methods and approaches to extract this information, in order to use them. Search engines [1] use these methods or approaches to extract and index the information contained in the web pages. Users use search engines to find useful information about the products they need, which shows the importance of search engines and having to equip them with good extraction methods to respond more accurately and in a relevant way to the need of users. Most of these search engines are based on keywords [2] to extract and index data from web pages, which explains the quality of the search results [3] of these engines which often return results that does not match the search performed, the result is not always relevant, hence the approach proposed in this article, it is a new approach that consists of linking the CSS incorporated on the e-commerce web page and GOODRELATIONS ontology used to index these web pages by means of a database, and from these CSS classes, generate a Wrapper to extract all the information about the products, which allow us to know the attribute of each product, it corresponds to which attribute of the ontology, i.e. its semantics, this will improve the relevance of the results of the research and respond more precisely to the need of the user.
关键词：Web Semantic; Information Extraction; Information Retrieval System; Semantic Indexing; E-Commerce; Ontology; GOODRELATIONS
