Scientific Formulas Extraction Oriented Towards Web Summarizing File Cards

Abstract:- This paper is devoted to the results of the Filimage system using Internet oriented technologies. Our solution is applied to textual documents and the main objective is to provide efficiency the e-document extraction containing mathematical objects. This work deals with the problem of the scientific formulas and the process designed to enable a semantic filtering. The significant interest of the adapted strategy is to bring out the multimedia contents use to perform an information retrieval. We introduce a description of the Contextual exploration method which is carried out for the text semantic filtering. Based on the access to user-relevant information, the system lies on the Natural Language Processing (NLP) to identify the visual components (formulas, tables, diagrams, images...) and the textual extracts at the same time which be used as their comments by A.M.A (Automatic Matching and Associating) operation. The paper gives an overview of the development and shows how is implemented the system. In order to illustrate the results of this process, we expose them by a visualization example containing automatic extracted scientific formulas. At last, we present how the extracted information are stored (scientific formulas, textual comments, title…) for a subsequent using of the results