Re: typical distribution of arities?

Hans Chalupsky <[email protected]> Mon, 23 Jul 2007 12:54:14 -0700
Newsgroups gmane.comp.ai.powerloom
Message-ID <[email protected]>
Rich,

here are a couple of datapoints:

PowerLoom's built-in ontology (see queries below):

66 Concepts
20 Unary relations
81 Binary relations
38 Ternary relations
 4 4-ary relations
 1 5-ary relation
10 variable-arity relations

Research Cyc 1.0:

3000000+ concepts
26000+ relations
 3438 unary relations
11935 binary relations
 3337 ternary relations
 1055 4-ary relations
  545 5-ary relations
   75 variable-arity relations

As you can see the Cyc relation numbers don't quite add up, since I
didn't run any queries but simply approximated the numbers by looking
at the relation extents of the various argNisa predicates (which
doesn't account for multiple or missing type assertions, etc.)  But it
should give you an approximate idea of the distribution.

Hans


STELLA(23): (retrieve all (concept ?r))
There are 66 solutions:
  #1: ?R=/TIMEPOINT-KB/DAY-OF-WEEK
  #2: ?R=/TIMEPOINT-KB/TIMEPOINT
  #3: ?R=/UNIT-KB/MASS-MEASURE
  #4: ?R=/UNIT-KB/DISTANCE-MEASURE
  #5: ?R=/UNIT-KB/AREA-MEASURE
  #6: ?R=/UNIT-KB/VOLUME-MEASURE
  #7: ?R=/UNIT-KB/TIME-MEASURE
  #8: ?R=/UNIT-KB/VELOCITY-MEASURE
  #9: ?R=/UNIT-KB/ACCELERATION-MEASURE
  #10: ?R=/UNIT-KB/ANGLE-MEASURE
  #11: ?R=/UNIT-KB/SOLID-ANGLE-MEASURE
  #12: ?R=/UNIT-KB/AMOUNT-OF-SUBSTANCE-MEASURE
  #13: ?R=/UNIT-KB/MOLALITY-MEASURE
  #14: ?R=/UNIT-KB/CONCENTRATION-OF-SUBSTANCE-MEASURE
  #15: ?R=/UNIT-KB/MOLAR-MASS-MEASURE
  #16: ?R=/UNIT-KB/FORCE-MEASURE
  #17: ?R=/UNIT-KB/PRESSURE-MEASURE
  #18: ?R=/UNIT-KB/DENSITY-MEASURE
  #19: ?R=/UNIT-KB/POWER-MEASURE
  #20: ?R=/UNIT-KB/WORK-OR-TORQUE-MEASURE
  #21: ?R=/UNIT-KB/REVOLUTION-MEASURE
  #22: ?R=/UNIT-KB/FREQUENCY-MEASURE
  #23: ?R=/UNIT-KB/ELECTRIC-CURRENT-MEASURE
  #24: ?R=/UNIT-KB/ELECTRIC-CHARGE-MEASURE
  #25: ?R=/UNIT-KB/ELECTRIC-POTENTIAL-MEASURE
  #26: ?R=/UNIT-KB/ELECTRIC-RESISTANCE-MEASURE
  #27: ?R=/UNIT-KB/ELECTRIC-CONDUCTANCE-MEASURE
  #28: ?R=/UNIT-KB/CAPACITANCE-MEASURE
  #29: ?R=/UNIT-KB/INDUCTANCE-MEASURE
  #30: ?R=/UNIT-KB/MAGNETIC-FLUX-MEASURE
  #31: ?R=/UNIT-KB/MAGNETIC-FLUX-DENSITY-MEASURE
  #32: ?R=/UNIT-KB/LUMINOUS-INTENSITY-MEASURE
  #33: ?R=/UNIT-KB/ILLUMINANCE-MEASURE
  #34: ?R=/UNIT-KB/DATA-MEASURE
  #35: ?R=/UNIT-KB/DIMENSIONLESS-MEASURE
  #36: ?R=/UNIT-KB/BASE-MEASURE
  #37: ?R=/UNIT-KB/MEASURE
  #38: ?R=FLOAT
  #39: ?R=MODULE
  #40: ?R=USER-THING
  #41: ?R=INTERVAL-CACHE
  #42: ?R=INTEGER
  #43: ?R=PROPOSITION
  #44: ?R=CONTEXT
  #45: ?R=STRING
  #46: ?R=BINARY-RELATION
  #47: ?R=FRAME-PREDICATE
  #48: ?R=SKOLEM-FUNCTION
  #49: ?R=COMPUTED-PROCEDURE
  #50: ?R=INTEGER-INTERVAL
  #51: ?R=THING
  #52: ?R=SCALAR-INTERVAL
  #53: ?R=NUMBER
  #54: ?R=LITERAL
  #55: ?R=SCALAR
  #56: ?R=STUFF
  #57: ?R=GROUP
  #58: ?R=LIST
  #59: ?R=NUMERIC-SET
  #60: ?R=SET
  #61: ?R=DUPLICATE-FREE-COLLECTION
  #62: ?R=COLLECTION
  #63: ?R=AGGREGATE
  #64: ?R=FUNCTION
  #65: ?R=RELATION
  #66: ?R=CONCEPT
STELLA(24): (retrieve all (and (arity ?r 1)
                               (fail (concept ?r))))
There are 20 solutions:
  #1: ?R=VARIABLE-ARITY
  #2: ?R=HIDDEN-RELATION
  #3: ?R=POLYMORPHIC
  #4: ?R=TOTAL
  #5: ?R=HANDLES-REVERSE-POLARITY
  #6: ?R=CLOSED
  #7: ?R=SINGLE-VALUED
  #8: ?R=MUTUALLY-DISJOINT-COLLECTION
  #9: ?R=HAS-PARTITION-MEMBERSHIP
  #10: ?R=EMPTY
  #11: ?R=COMPUTED
  #12: ?R=ABSTRACT
  #13: ?R=DUPLICATE-FREE
  #14: ?R=ORDERED
  #15: ?R=COMMUTATIVE
  #16: ?R=REFLEXIVE
  #17: ?R=IRREFLEXIVE
  #18: ?R=SYMMETRIC
  #19: ?R=ANTISYMMETRIC
  #20: ?R=TRANSITIVE
STELLA(25): (retrieve all (arity ?r 2))
There are 81 solutions:
  #1: ?R=ARITY
  #2: ?R=INVERSE
  #3: ?R=subset-of
  #4: ?R=RELATION-COMPUTATION
  #5: ?R=UPDATE-PROPOSITION-DEMON
  #6: ?R=GOES-TRUE-DEMON
  #7: ?R=GOES-FALSE-DEMON
  #8: ?R=GOES-UNKNOWN-DEMON
  #9: ?R=>=
  #10: ?R=MEMBER-OF
  #11: ?R=INEQUALITY
  #12: ?R=PHRASE
  #13: ?R=DISJOINT
  #14: ?R=CONCEPT-PROTOTYPE
  #15: ?R=RELATION-SPECIALIST
  #16: ?R=RELATION-CONSTRAINT
  #17: ?R=RELATION-EVALUATOR
  #18: ?R=CARDINALITY
  #19: ?R=DIRECT-SUBRELATION
  #20: ?R=DIRECT-SUPERRELATION
  #21: ?R=SUBRELATION
  #22: ?R=SUPERRELATION
  #23: ?R=PROPER-SUBRELATION
  #24: ?R=PROPER-SUPERRELATION
  #25: ?R=EQUIVALENT-RELATION
  #26: ?R=SYNONYM
  #27: ?R=OBSERVED-CARDINALITY-OF
  #28: ?R=INSTANCE-OF
  #29: ?R=INTERVAL-CACHE-OF
  #30: ?R=>
  #31: ?R=<
  #32: ?R==<
  #33: ?R=LEXEME
  #34: ?R=COMMENT
  #35: ?R=DOCUMENTATION
  #36: ?R=ISSUE
  #37: ?R=EXAMPLE
  #38: ?R=IMAGE-URL
  #39: ?R=IST
  #40: ?R=TYPE-OF
  #41: ?R=PORTION-OF
  #42: ?R=DOMAIN
  #43: ?R=RANGE
  #44: ?R=NUMERIC-INTERVAL
  #45: ?R=SQRT
  #46: ?R=SQUARE-ROOT
  #47: ?R=ABSOLUTE-VALUE
  #48: ?R=NEGATE
  #49: ?R=COLLECT-INTO-SET
  #50: ?R=LENGTH-OF-LIST
  #51: ?R=MINIMUM-VALUE
  #52: ?R=MAXIMUM-VALUE
  #53: ?R=SUM
  #54: ?R=MEAN-VALUE
  #55: ?R=MEDIAN-VALUE
  #56: ?R=STANDARD-DEVIATION
  #57: ?R=VARIANCE
  #58: ?R=LENGTH
  #59: ?R=OBJECT-NAME
  #60: ?R=NAME-TO-OBJECT
  #61: ?R=INTEGER-INTERVAL.INTERVAL-LOWER-BOUND
  #62: ?R=INTEGER-INTERVAL.INTERVAL-UPPER-BOUND
  #63: ?R=CASE-MATCH
  #64: ?R=SOFT-EQ
  #65: ?R=CONTEXT-OF
  #66: ?R=COVERING
  #67: ?R=DISJOINT-COVERING
  #68: ?R=CHEAP-DISJOINT
  #69: ?R=REFUTATION-DISJOINT
  #70: ?R=/UNIT-KB/NUMERATOR-MEASURES
  #71: ?R=/UNIT-KB/DENOMINATOR-MEASURES
  #72: ?R=/TIMEPOINT-KB/TIMEPOINT-OF
  #73: ?R=/TIMEPOINT-KB/DURATION-OF
  #74: ?R=/UNIT-KB/MEASURE-UNIT
  #75: ?R=/UNIT-KB/VALUE-MEASURE
  #76: ?R=/UNIT-KB/COMENSURATE-UNITS
  #77: ?R=/UNIT-KB/U-ABS
  #78: ?R=/UNIT-KB/U-SIGNUM
  #79: ?R=/UNIT-KB/U-MINIMUM-VALUE
  #80: ?R=/UNIT-KB/U-MAXIMUM-VALUE
  #81: ?R=/UNIT-KB/U-SUM
STELLA(26): (retrieve all (arity ?r 3))
There are 38 solutions:
  #1: ?R=FORK
  #2: ?R=PARTITION-MEMBERSHIP
  #3: ?R=THE-ROLESET
  #4: ?R=RANGE-MIN-CARDINALITY
  #5: ?R=RANGE-MAX-CARDINALITY
  #6: ?R=RANGE-TYPE
  #7: ?R=NTH-DOMAIN
  #8: ?R=FILLERS
  #9: ?R=RANGE-CARDINALITY-LOWER-BOUND
  #10: ?R=RANGE-CARDINALITY-UPPER-BOUND
  #11: ?R=RANGE-CARDINALITY
  #12: ?R=NUMERIC-MINIMUM
  #13: ?R=NUMERIC-MAXIMUM
  #14: ?R=UNION-OF
  #15: ?R=+
  #16: ?R=-
  #17: ?R=*
  #18: ?R=/
  #19: ?R=PROJECT-COLUMN
  #20: ?R=STRING-CONCATENATE
  #21: ?R=/UNIT-KB/UNITS
  #22: ?R=/TIMEPOINT-KB/TIMEPOINT-YEAR
  #23: ?R=/TIMEPOINT-KB/TIMEPOINT-MONTH
  #24: ?R=/TIMEPOINT-KB/TIMEPOINT-DAY
  #25: ?R=/TIMEPOINT-KB/TIMEPOINT-DAY-OF-WEEK
  #26: ?R=/TIMEPOINT-KB/TIMEPOINT-HOUR
  #27: ?R=/TIMEPOINT-KB/TIMEPOINT-MINUTE
  #28: ?R=/TIMEPOINT-KB/TIMEPOINT-SECOND
  #29: ?R=/TIMEPOINT-KB/TIMEPOINT-DATE
  #30: ?R=/TIMEPOINT-KB/TIMEPOINT-TIME
  #31: ?R=/UNIT-KB/U+
  #32: ?R=/UNIT-KB/U-
  #33: ?R=/UNIT-KB/U*
  #34: ?R=/UNIT-KB/U-DIV
  #35: ?R=/UNIT-KB/U-WITHIN-DELTA
  #36: ?R=/UNIT-KB/U-WITHIN-FACTOR
  #37: ?R=/TIMEPOINT-KB/TIME-
  #38: ?R=/TIMEPOINT-KB/TIME+
STELLA(27): (retrieve all (arity ?r 4))
There are 4 solutions:
  #1: ?R=QUALIFIED-RANGE-MIN-CARDINALITY
  #2: ?R=QUALIFIED-RANGE-MAX-CARDINALITY
  #3: ?R=QUALIFIED-RANGE-CARDINALITY
  #4: ?R=SUBSTRING
STELLA(28): (retrieve all (arity ?r 5))
There is 1 solution:
  #1: ?R=STRING-MATCH
STELLA(29): (retrieve all (arity ?r 6))
No solutions.
STELLA(30): (retrieve all (variable-arity ?r))
There are 10 solutions:
  #1: ?R=DIFFERENT
  #2: ?R=QUERY
  #3: ?R=BOUND-VARIABLES
  #4: ?R=CUT
  #5: ?R=VALUE
  #6: ?R=HOLDS
  #7: ?R=LISTOF
  #8: ?R=SETOF
  #9: ?R=COLLECTIONOF
  #10: ?R=SKOLEM-FN
STELLA(31): 

--------------------------------------------------------------------------
Hans Chalupsky, PhD                     USC Information Sciences Institute
Project Leader, Loom KR&R Group         4676 Admiralty Way
<[email protected]>                          Marina del Rey, CA 90292
(310) 448-8745
--------------------------------------------------------------------------

>>>>> Rich Morin <[email protected]> writes:

> I'd like to get an idea of the distribution of N in N-ary
> relationships, as expressed in knowledge bases, etc.  That
> is, what percentage of relationships can be expressed as
> binary, trinary, etc.

> If anyone has a suitable test case and would be willing to
> run a quick tally, I'd love to see the results.  I'd also
> be happy to receive anecdotal impressions, etc.

> Feel free to send the information to me off-list...

> -r
> -- 
> http://www.cfcl.com/rdm            Rich Morin
> http://www.cfcl.com/rdm/resume     [email protected]
> http://www.cfcl.com/rdm/weblog     +1 650-873-7841

> Technical editing and writing, programming, and web development
> _______________________________________________
> powerloom-forum mailing list
> [email protected]
> http://mailman.isi.edu/mailman/listinfo/powerloom-forum