TactiNET

Abstract

In order to overcome the difficulties of interaction with digital documents, the research can be conducted in two main directions: (1) design or improve techniques which make reading or entering data in digital documents easier, or (2) modify the document to make it more accessible using existing tools. In the context of non-visual access to digital documents, these directions, or combining them, have caused a significant scientific production. The first step is to design sensory substitution devices [1]. The second step, which could be involved in the areas of Natural Language Processing (NLP) and Information Retrieval (IR), is transforming the documents to adapt them with the constraints of a non-visual interaction (oral, touch or multimodal). We propose a solution that combines these two steps to solve a main difficulty encountered by blind persons when browsing the Web on touch-screen devices: the lack of overall perception of Web pages. First experiments show promising results in the paradigm "design for all".

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 1 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.