FuXi 0.95b.dev-r229 released to Cheeseshop and a User Manual / Wiki

FuXi has taken the first of a few remaining final steps towards a 1.0
release candidate.  A handful of updates have been committed to the
subversion repository and released (0.95b) to Cheeseshop

So, you can install it via:

  easy_install FuXi

In addition, the beginnings of a user manual has been added to the
Google Code wiki:

http://code.google.com/p/python-dlp/wiki/FuXiUserManual

The collective changelog for the most recent updates is below.  At
this point, the DLP and RETE-UL mechanisms have gone through
significant testing.  The module that is the least mature (and will be
the focus leading up to a 1.0 release candidate) will be Proof
generation.  I'm considering implementing Magic Sets [1] to improve
the naive proof generation mechanism which currently just forward-
fires until the goal is reached:

"One method, called “magic sets,” is a general algorithm for rewriting
logical rules so that they may be implemented bottom-UP (= forward
chaining) in a way that cuts down on the irrelevant facts that are
generated. The advantage of this scheme is that by working bottom-up,
we can take
advantage of efficient methods for doing massive joins."

[1] http://doi.acm.org/10.1145/6012.15399

------>%---------------------

- parsing of additional rulesets to the network (for after-
construction-parsing). Fixed built-in interfaces.  .etc.
- Support for non denoting builtins (users can specify their own
builtin predicates)
- Make Clauses hashable (useful for identifying duplicates in a
ruleset)
- Added utility method for generating DLP mechanism on a network
instance
- fixed Network.reset to also reset instanciation statistics
- fixed serialization of 'that' keyboard with owl:intersectionOf
- added method for clearing boolean classes
- added checks for rule (and terminal node) duplication
- Fixed DLP-based generation of owl:intersectionOf.  Conjunctions of
existential restrictions are now flattened bypassing redundant
intermediate BNode classifications.
- Added --pDSemantics options to command line (parses pD ruleset into
network)
- Removed minor redundancies
- Added owl:hasValue transformation to DLP module (so OWL graphs with
owl:hasValue will generate customized rulesets rather than use the
generic owl:hasValue rule
- Fixed handling of disjunction in a rule head (skips the rule
completely)
- Compressed the generation of rules from owl:intersectionOf (no
longer
generates redundant rules for the intermediate anonymous classes)
- graceful failure of pydot import

Received on Thursday, 31 July 2008 21:49:35 UTC