Hi, my name is Patrick van Bergen.
This website is my attempt to answer the following question:
What does it take to make a good Natural Language Interface?
To answer this question I examine historic NLI systems and extract their features. Each of these systems is special and highlights a different aspect of the task.
The website allows you to view features per system module, to see which features are more common than others, to select all systems that implement a given feature, and to compare the features of two or more systems.
Along the way, I try to explain these features as well as I can, and provide examples and links to more information.
I hope this will give you a better insight into the nature of NLI systems.
I just include the systems that make a true effort to allow the user to approach a database with the full range of natural language.
Further restrictions I make:
I might even include a system that fulfills only a single one of these requirements, but it would have to be a historically significant contribution.
I am just a programmer interested in Natural Language. I'm not a scientist. On this website I try to be as scientific as I can, but the fact is that many of the features and categories presented on this website were made up by me as I tried to chart the territories of Natural Language Interaction. Others may not share my interpretation, and would make other classifications. Take this website as a first draft.
When I was creating this website I found much inspiration from the the following overview article. I may quote from it from time to time.
Natural Language Interfaces to Databases - An Introduction (1995): I. Androutsopoulos, G.D. Ritchie, P. Thanisch