An ambiguous tag-based query reformulation technique for an effective semantic-based social image research

The Tag-based Document Retrieval technique was adopted for long time as an intuitive way to search for images shared on social networks. Nevertheless, the tag-based queries are often too ambiguous, and consequently they do not constitute an efficient solution for retrieving the most relevant images that meet the users’ needs. As an alternative, the Semantic-based Social Image Retrieval technique has emerged. The policy of this technique consists in retrieving the relevant images covering as much possible the topics that a given ambiguous query may have. In this paper, we propose a novel technique at the ambiguous query pre-processing level, which aims at moving from an ambiguous tag-based query towards a semantic-based one, by relying on a set of predefined ontological semantic rules. Thorough experiments using 8 ambiguous queries over a collection of 25.000 socio-tagged images shared on Flickr service prove the effectiveness of our technique.