In this paper we introduce an algorithmic approach, capable of creating a semantic network with concatenated terms and phrases (hash tags) from collectively postings on the Twitter sphere. This network could be exploited for query expansion provision in respect to users¢ information needs, without considering any other prior knowledge or access in search logs or browsing history records. For evaluation purposes, we compare our query expansion approach algorithm with query suggestions provided by well-known search engines and mainstream media services (e.g. Google, Yahoo!, Bing, NBC and Reuters), as well as by enrolling a team of human editors, who provided subjective comparisons in respect to the Google Hot Searches service. The results are quite promising, showing that our proposal semantically expands the user¢s initial query with related terms adapted to social trends and knowledge.