From Archives Management to Knowledge Extraction: Tools and Impact Evaluation
Résumé
The aim of th is paper is to illustrate the application of some automatic knowledge extraction tools on archive records. These tools are based on Machine and Deep Learning technologies that facilitate the implementation process and allow the creation of scalable solutions. The purpose of the work is to automatically extract metadata, keywords, terms, phrases, entities, tables and graphs from documents which are characterized by complex informational dimensions. The ultimate goal is to measure the quality of the extraction obtained. In particular, the work aims to evaluate the results obtained from the application of the selected tools on a specific corpus of archival documents. The evaluation process will pay particular attention to the accuracy of the results on the semantic level.