C-Phrase: A natural language interface to databases under BSD

Dear All,<p>there is a chance that some of you may remember me from before: <a href="http://news.ycombinator.com/item?id=275973" rel="nofollow">http://news.ycombinator.com/item?id=275973</a><p>It's a long story, but the short of it is that circumstances forced me to follow option #2. I have open sourced the system. It's under BSD and in LISP.<p>I have not been so involved in c-phrase lately, but that is about to change in the new year -- I will have a reduced teaching commitment. :-)<p>Any ways I revised the installation guide recently and was wondering if any one could verify that they are able to install the system. Any questions or feedback on the system/code/design would be greatly appreciated.<p>Regards,<p>Michael Minock <a href="http://www.cs.umu.se/~mjm" rel="nofollow">http://www.cs.umu.se/~mjm</a>

Quite interesting...<p>I have a fear of human-language-like programming languages, after working on a Progress database. Too many keywords, too easy to forget how to do specific things, etc. Of course, that's a very basic one.<p>This looks like it'd be a fantastic <i>alternate</i> database accessor, precisely for custom queries like the video shows. I'd hate to be a DBA trying to optimize things though, it seems the higher abstraction away from the actual query means optimizations would occasionally require unnatural query text.<p>All that said, I'm definitely going to have to look over the code. Thanks for open-sourcing it!

&#62; I will have a reduced teaching commitment. :-)<p>I found your classes quite engaging. So the "reduced-teaching-:-)" might not be an attribute of the student community.<p>Still, good luck in further developing C-Phrase. I remember I was quite impressed by its capabilities (when you had me testdrive it, little less than two years ago).<p>From what I remember, it required a person with nonzero understanding and intelligence (call it an 'expert') to set up and tune the ontologies (ongoingly). But at the same time the natural language interface could be used by laymen, with the expert working the system behind the scenes, ongoingly extending it wherever needed (maybe driven by failed queries that look right, but don't work). How much skill do you think such an expert needs to possess? Or do you plan on cutting out this 'middle man'?<p>PS The video (<a href="http://www.youtube.com/watch?v=fWio8bHq4wQ" rel="nofollow">http://www.youtube.com/watch?v=fWio8bHq4wQ</a>) at 2:20 gave me a laugh because the guy seems to attempt middle-mouse-button-pasting an X11 selection on a system that's not X11. Quite some people here will know how that feels ;-)

Some examples of what this can do would be nice. (No, I'm not watching a video, sorry.)<p>At first glance, the manual <i>requires</i> knowledge of Cobb's notations - try to make this unnecessary, as many systems administrators won't know it. (I'm in academia, but my degree is in Mathematics instead of computer science; the notation <i>looks</i> trivial, but it's one more obstacle.)<p>In short, the project looks interesting, but it was hard to evaluate at a glance.