An Overview of NLIDB Approaches and Implementation for Airline Reservation System

Abstract

Relational databases are queried using database query languages such as SQL. Natural language interfaces to databases (NLIDB) are systems that translate a natural language sentence into a database query. In this modern techno-crazy world, as more and more laymen access various systems and applications through their smart phones and tablets, the need for Natural Language Interfaces (NLIs) has increased manifold. The challenges in Natural language Query processing are interpreting the sentence correctly, removal of various ambiguity and mapping to the appropriate context. Natural language access problem is actually composed of two stages-Linguistic processing and Database processing. NLIDB techniques encompass a wide variety of approaches. The approaches include traditional methods such as Pattern Matching, Syntactic Parsing and Semantic Grammar to modern systems such as Intermediate Query Generation, Machine Learning and Ontologies. In this report, various approaches to build NLIDB systems have been analyzed and compared along with their advantages, disadvantages and application areas. Also, a natural language interface to a flight reservation system has been implemented comprising of flight and booking inquiry systems.