Course talk:CPSC522/Inductive Logic Programming

Hi Ritika,

This page is clear and interesting, the introduction and background part is perfect, but I have several suggestions:

1. At the "Problem definition" section, you first time reference the "CHILL", but at that time I really don't know what is "CHILL", I suggest move the "Overview of CHILL" in front of the "Problem definition", or tell us a little more what is "CHILL" before introduce the problem.

2. I think you are comparing the "CHILL" whith the "Geobase" (I am not sure), I saw the figure of the experiment result, but I cant find the reason why CHILL did better than Geobase. I think it would be better to simply explain that.

Sincerely,

Junyuan Zheng

Hi Junyuan, Thank you so much for your feedback. 1. I should certainly look into describing CHILL before the problem definition. 2. Yes the paper compares CHILL with Geobase which is an already existing natural language database interface system. I shall elaborate on the explanation of how CHILL performs better than Geobase below the accuracy graph. Thank you for your suggestions. :)

Warm regards,

Ritika

Hi Junyuan, I have made the suggested changes. Let me know if it makes sense now.

Thanks :)

Ritika

Hi Ritika,

Problems are solved! It looks perfect now!

Sincerely,

Junyuan Zheng

Hi Ritika
Nice page! The introduction and background helped me in understanding the page. Some questions I had :

1. Is CHILL accessible to us?
2. Is there any performance evaluation comparison for ILP and CHILL?
3. Considering the fact that the papers were written in 1988 and 1996 have there been any improvements made to the systems?

Hi Samprity, Thank you for your feedback. 1. Yes CHILL is accessible to us. I shall add links and some background on how we can use it. 2. So CHILL uses ILP in its functioning. CHILL is like a parser acquisition system which employs Inductive Logic Programming. So I guess we could say that a good performance evaluation of CHILL gives us faith in ILP. 3. I should include insight into some recent projects with CHILL. Thank you so much for your insightful comments :) Ritika

Hi Samprity, I've made the suggested changes. Let me know if it makes sense now.

Thanks :)

Ritika

Hi Ritika, Well written page with nice examples. I have the following feedback 1. Is it possible to give a little bit of discussion of how the algorithms work, specially after the pseudo-codes? 2. A few more external links for the naive user would make it even better in my opinion. Hope this helps :)

Hi Abed, Thank you for your feedback and suggestions. 1. I would elaborate on the pseudo-code and try my best to explain it further. 2. And yes, I think I should add some more information on the recent projects using CHILL and more external links for the naive user. Thanks a lot.  :)

Ritika

Hi Abed, I've made the suggested changes. Elaborated upon the algorithm and added further resources/recent work section. Let me know if this makes sense.

Thanks :)

Ritika