Towards Linked Data Extraction From Tweets

International audienceMillions of Twitter users post messages every day to communicate with other users in real time information about events that occur in their environment. Most of the studies on the content of tweets have focused on the detection of emerging topics. However, to the best of our knowledge, no approach has been proposed to create a knowledge base and enrich it automatically with information coming from tweets. The solution that we propose is composed of four main phases: topic identification, tweets classification, automatic summariza-tion and creation of an RDF triplestore. The proposed approach is implemented in a system covering the entire sequence of processing steps from the collection of tweets written in English language (based on both trusted and crowd sources) to the creation of an RDF dataset anchored in DBpedia's namespace