ABSTRACT: The web contains very large amount of unstructured text. The process of converting unstructured text to structured with semantic information annotation can provide useful summaries for both humans and machines. The semantic relation is one of the most important parts of semantic information. Hence, extracting semantic relations held between entities in a text is important in many natural language understanding applications like question answering, conversational agents, summarization. There are several proposed methods in this area. Hand built patterns, bootstrapping methods, supervised methods, unsupervised methods, and Distant Supervision (DS) are examples. The purpose of this work is to review various methods used for Relation Extraction (RE). For each approach, respective motivation is discussed and the merits and demerits compared. Discussion about various datasets is also included in this paper.