Unified Semantic Space for a Novel Multimodal Approach to Document Similarity

Simone Scannapieco, Andrea Ponza, Claudio Tomazzoli · 2021 IEEE 6th International Forum on Research and Technology for Society and Industry (RTSI) · 2021

The paper explores a new method to represent digital documents - containing several sources of information (text, images, videos, multimedia galleries) - in a semantic space able to elevate the importance of each informative channel in giving a deeper meaning to the document. The proposed approach finds several applications such as Web Content Mining, semantic classification and document topic extraction.

Read the paper · More papers on PaperTik