In these days, much of the information published on the Web is published on social media, represented through social networks such as Facebook or Twitter, to name only the most prominent.Each of the media and social networks has its own scheme of operation and different working characteristics; for example, Twitter is a social network where millions of daily messages called tweets are exchanged. The message labels, called hashtags, can be used to identify the subject of the message. The message may also include links to other resources that expand the original content or show interesting information. Another kind of information present in Twitter is the relationships between users, the most common of which is a non-reciprocal relationship named "Following".The scope of this paper is to use information that is published on Twitter to extract and recommend Open Educational Resources, which will be used in the StartUP project. The
StartUP project (Intelligent training needs assessment and Open Educational Resources to foster Entrepreneurship) is co-funded with support by the European Commission under the LifelongLearning Programme, which has as specific objective to provide effective open educational resources corresponding to individual learning needs.The extraction of information posted on social networks is solved in this research through the use of linked data, that allow retrieving resources and link them with other external sources, graphs that help represent the working scheme of a social network, and with social network analysis (SNA), a technique used to discover relevant information that goes beyond individual properties.The results obtained are a set of recommendations about users (identified as experts), hashtags (thematically related), and URLs (digital resources), according to the main competence areas defined by StartUp. This information will form part of learning paths provided by the project platform.