In this paper we present an enhanced algorithm with modified approach to extricate various Triplets i.e. subject-predicate-object from Natural language sentences. The Treebank Structure and the Typed Dependencies obtained from Stanford Parser are used to elicit multiple triplets from English Sentences. Typed Dependencies represents grammatical connections among the words of any sentence and represents how triplets are associated. The intended interpretation behind the extraction of Triplets is that the subject is acting on the object in a way described by the predicate. In graphical form it can be considered that subject and object will be acting as nodes i.e. entities and predicate as edges i.e. relationship. The resulting triplets and relations can be useful for building and analysis of a social network graph and for generating communication pattern and Information retrieval.