Entity Linking (EL) is a critical task in Information Extraction (IE) that involves associating Named Entities (NEs) mentioned in text with their corresponding entity in a Knowledge Base (KB). This intersection of Natural Language Processing (NLP) and Semantic Web Exploitation (SWE) has mainly been investigated on large texts, such as the newswire. However, the shift towards analysing microposts - short, informal social media content - has revealed significant shortcomings in the performance of conventional EL methods. We introduce UFEL, a novel zero-shot strategy for EL task in the context of microposts conversations. Our methodology capitalises on open Semantic Web resources, including the Wikipedia and Wikidata APIs, DBPedia and Wikidata as KBs, and operates under the assumption that consecutive NEs within a micropost exhibit a high degree of semantic interconnection. Furthermore, the simplicity of our scoring mechanism makes our solution efficient and easily understandable while reaching 75% F1-score, setting a new state of the art in the context of French microposts data. We evaluate UFEL with respect to three other systems, including ReFinED, a deep learning system from Amazon.