Natural Language Database Interface System

AKA: Natural Language Access to Database, NLIDB, NLDBI.
Context:
- It can be solved by a Natural Language Database Interface Task by Implementing Natural Language Databese Interface Algorithms.
- It can range from being a simple Natural Language Interface to Databases System, to being a Web-Based Natural Language Interface to Databases System, to being a Neural Natural Language Interface to Databases System.
Example(s):
- AlLaDIn,
- DBPal,
- ENGLISH WIZARD,
- INTELLECT,
- LANGUAGEACCESS,
- LOQUI.
- PARLANCE,
- PRECISE,
- NaLIX,
- NaLIR,
- NATURAL LANGUAGE NLIDB.
- …
Counter-Example(s):
- a Natural Language Interface to Ontologies System such as:
  - PANTO;
  - Querix;
  - Aqualog;
  - Ginseng;
- a Question Answering System such as:
  - BASEBALL,
  - EAGLi,
  - ELIZA,
  - LUNAR,
  - SHRDLU.
See: User Interface, Natural Language Processing, Natural Language Understanding, Natural Language Generation, Question Answering Task.

References

(Utama et al., 2018) ⇒ Prasetya Utama, Nathaniel Weir, Fuat Basik, Carsten Binnig, Ugur Cetintemel, Benjamin Hattasch, Amir Ilkhechi, Shekar Ramaswamy, and Arif Usta. (2018). “An End-to-end Neural Natural Language Interface for Databases.” In: CoRR Journal, abs/1804.00401. arXiv:1804.00401
- QUOTE: Structured query language (SQL), despite its expressiveness, may hinder users with little or no relational database knowledge from exploring and making use of the data stored in an RDBMS. In order to effectively leverage their data sets, users are required to have prior knowledge about the schema information of their database, such as table names, columns and relations, as well as a working understanding of the syntax and semantics of SQL. These requirements set “a high bar for entry” for democratized data exploration and thus have triggered new research efforts no develop alternative interfaces that allow non-technical users to explore and interact with their data conveniently. While visual data exploration tools have recently gained significant attention, Natural Language Interfaces to Databases (NLIDBs) appear as highly promising alternatives because they enable users to pose complex ad-hoc questions in a concise and convenient manner.

(Li & Jagadish, 2014) ⇒ Fei Li, and H. V. Jagadish. (2014). “Constructing An Interactive Natural Language Interface for Relational Databases.” In: Proceedings of the VLDB Endowment Journal, 8(1). doi:10.14778/2735461.2735468
- QUOTE: NLIDBs have many advantages over other widely accepted query interfaces (keyword-based search, form-based interface, and visual query builder). For example, a typical NLIDB would enable naive users to specify complex, ad-hoc query intent without training. In contrast, keywords are insufficient to convey complex query intent, form-based interfaces can be used only when queries are predictable and limited to the encoded logic, and visual query builders still requires extensive schema knowledge of the user.
  (...) we propose an NLIDB comprising three main components: a first component that transforms a natural language query to a query tree, a second component that verifies the transformation interactively with the user, and a third component that translates the query tree into a SQL statement. We have constructed such an NLIDB, and we call it a NaLIR (Natural Language Interface to Relational databases).