Minimizing human intervention for NLI to relational database

In order to access data in database, we utilize database query language such as structure query language (SQL). However, these languages are complex and difficult for users not to be familiar with database. Natural Language Interfaces for Database system (NLIDBs) is research community to study natural language interfaces for database instead of database queries. Although NLIDBs has this advantage, it has not gained the expected rapid and wide commercial acceptance. The reasons are that it is difficulty to process linguistic problems such as conjunction/disjunction, negation, quantifier scooping, and anaphora and that it is a burden to human load for transport to a new domain. This book is focused on portability and proposes concept model which can adapt to a new domain with a little human intervention. Although a question with same intention, users can use various linguistic expressions. And database can be also designed variously in a domain. The purpose of intermediate representation as concept model is to narrow the gap between the linguistic expressions and database query languages in database. In this book, a approach to realize concept model is to use semantic dictionary