Automated Generation of E-R Diagram from a Given Text in Natural Language
Sutirtha Ghosh, Prasenjit K. Mukherjee, Baisakhi Chakraborty, Rezaul Bashar · 2018
This paper proposes an Automated E-R Diagram Generation (AGER) System that can generate E-R Diagram from a given text in Natural Language given as input. Natural language texts are used to perform the information extraction by parsing the syntax of the sentences and semantically analyzing their content. The AGER System parses an input text in natural language, semantically analyzes it and internally uses some domain specific database and POS tagging to detect Entity and Relations from the given passage and builds a graph that represents the E-R Diagram. The E-R Diagram can be traversed to generate Data Definition Language to create the actual relations in any RDBMS system. The abstract nature of entity relationship diagram makes it difficult for database designers to directly create E-R Diagram from the natural language text input statement which is used to create the physical model of the database. In the requirement analysis phase of any software design, often Databases designers need to have elaborate discussions with the client on the use cases of the databases and at the end of the work they come up with a neat E-R Diagram which shall be used in subsequent phases to physically realize the relations and implement them. The AGER system proposed here is aimed to assist database designers to create E-R Diagram directly from Client’s requirements in natural language.