Automatic annotation of images, pictures or videos comments for text mining guided by no textual data

Florence Le Priol

The Text mining guided by No Textual data (TNT) is not intended to extract the information contained in the images, aiming the information included in the text that describes these images. In other words, it aims to present to the reader the information about the images next to them, regardless of its real position in the document. Reading, while focusing on no textual data (images, pictures, videos, sounds, etc.), their caption and some comments identified in the text is faster than reading the entire text and gives a good summary of the document. We present in this article, a tool for automatic text mining guided by no textual data (fixed or moving images, photographs, video, sound ... ). This tool, based on contextual exploration method and EXCOM platform, can automatically annotate the information concerning no textual data, it can make the link between annotated text and corresponding no textual data and propose navigations among these textual and no textual information.

Subjects: 13. Natural Language Processing; 1.10 Information Retrieval

Submitted: Feb 25, 2008


This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.