home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.ai
- Path: sparky!uunet!cs.utexas.edu!zaphod.mps.ohio-state.edu!uwm.edu!linac!att!cbnewsk!cbnewsj!att-out!walter!bellcore.com!caseau
- From: caseau@bellcore.com (Yves J Caseau)
- Subject: Re: AI Tools
- Message-ID: <1993Jan6.174048.19263@walter.bellcore.com>
- Sender: news@walter.bellcore.com
- Nntp-Posting-Host: france.bellcore.com
- Organization: Bell Communications Research, Morristown, NJ
- References: <derder-060193094356@mac_rfmq-3.imag.fr>
- Date: Wed, 6 Jan 93 17:40:48 GMT
- Lines: 30
-
- A while ago, I posted a request for a series of short programs that I could use
- to benchmark inference engines. I received many answers from people that were
- also interested in collecting such a set of benchmarks, but very few programs.
-
- After looking at what I had, and removing similar things (such as all the possible
- versions of monkay & banana), I just have a set of four reasonably short programs
- that are representative of what people commonly do with an inference engine.
- It contains:
-
- title what from
- monkey A planner M. Porcheron
- queens A naive exhaustive queens R. Lissajoux
- (Gauss problem) solver
- triplet Deducing values of logical M. Wallace
- variables bound by OR triplets
- filter Data filtering J. Britt
-
- If you are interested, I can send you the files through e-mail, together with
- some preliminary performance results. If many persons are interested, I will
- make it available through ftp.
-
- I am aware that this benchmark set is crude, but it is small enough that porting
- a test case to any inference engine should be little work, which was the original
- goal. Also, looking at it carefully shows that it represents a good blend of
- various aspects of rule processing.
-
- Any (constructive) criticism (or even better, any addition) is (warmly) welcomed.
-
- -- Yves Caseau
- caseau@bellcore.com
-