VIPReader: A Light News Reader for the Visually Impaired Person

摘要

Nowadays, the whelming online news information often bewilders the readers, especially the visually impaired persons, who, instead of reading, listen to news articles line by line with the help of a screen reader. Therefore, to improve the news accessibility of these people, we develop a news reader (VIPReader) to assist them with two unique functions, i.e., Multi-document summarisation and ontology-based recommendation. In particular, the summarisation engine of VIPReader generates a summary across several relevant documents to assist a quick view for visually impaired people, and its ontology-based recommendation engine suggests relevant news to visually impaired readers based on concepts of news and user preference. Experiments on visually impaired people show the easy information accessibility of the proposed VIPReader.