Google Filters

Google search results

It is not unusual for Google searches to return 100,000 URLs. Google does not provide good tools for filtering its search results to find what you really want.

• no category information for URLs
• no search within current set of URLs

Google searches begin with category information. Originally, Google used the Open Directory Project hierarchy

	http://mKRmKE.org/knowledge/google/Directory.ho

	http://mKRmKE.org/knowledge/google/search.ho

The mKE "google" command returns a set of mKR propositions which describe the Google search results. For example,

	$ google search "squire crow count" >squire.crow.count.mkr

	http://mKRmKE.org/knowledge/rhm/theory/Google/squire.crow.count.mkr

mKR pipelines

A Google filter pipeline is invoked like this.

	$ ke -m -google
	...
	ke$ {
		! google search terms > terms.mkr done;
		do empty od proposition done;
		do read from terms.mkr done;
	    }
		| at view=result1; { ... }
		| at view=result2; { ... }
		| ...
		| at view=resultN; { ... };
	ke$ exit;
	$

Webpage properties

Google uses proprietary Webpage properties for searching.
I am currently exploring some open Webpage properties which may be useful as Google filters. For example.

• table of contents (mKE "toc" command)
$ toc http://mKRmKE.org/ <TITLE>mKE (my Knowledge Explorer) and the mKR language</TITLE> <h2> <h2> Screen Shot </h2> <H2> Who am I? </H2> <H2> Why did I invent mKR? </H2> <H2> What does "m" stand for? </H2> <H2> What is the mKR language? </H2> <H2> What is Knowledge Explorer? </H2> <H2> What is Knowledge? </H2> <H2> What is a KnowledgeBase? </H2> <H2> What is Context? </H2> <H2> Semantic Web Applications </H2> <H2> More Information </H2> <H3> The mKR language </H3> <H3> The mKE program </H3> <H3> Knowledge </H3> <H3> mKR/mKE documentation </H3> <H2> Download mKE and other useful programs </H2> <H3> Download Knowledge Explorer 8.0 now <H3> Download Unicon 11.0 if you want to create <H3> Download other useful programs if desired<BR></H3> $
• word bag (mKE "wordbag" command)

$ wordbag http://mKRmKE.org/kb/spo.rdf.html | sort -rn
24*#
13*iss
12*is
8*//
7*Class
6*Resource
6*Property
5*/
4*subClassOf
4*s
4*member
4*Container
4*::
3*type
3*subPropertyOf
3*ingroup
2*subject
2*relation
2*property
2*isu
2*isg
2*hierarchy
2*exgroup
2*cps
2*Statement
2*Seq
2*Bag
2*Alt
2*///
1*xmp
1*value.
1*value
1*statement
1*species
1*set
1*rel
1*rdfs:subPropertyOf
1*rdfs:subClassOf
1*rdf:Resource
1*rdf:Property
1*rdf:Class
1*qualifiers
1*object.
1*mKR
1*list
1*isany
1*isall
1*isa
1*individual
1*i:type
1*i:subPropertyOf
1*i:subClassOf
1*html
1*group
1*genus.
1*file
1*existent
1*enum
1*entity
1*end
1*class.
1*binary
1*begin
1*attribute
1*RDF
1*KEHOME/kb/spo.rdf.html
1*Jun/10/2008
1*Feb/22/2007
1*/xmp
1*/html
$ 

• Snippet

ke$ every p isu gpage; {$p has Snippet=?;};
gpage_1 has Snippet = "Squire Crow was born in Washington county, and was a farmer and
 carpenter, also 
  a justice of the peace a number of years. Mrs. Crow was born in Ohio.
...";
gpage_10 has Snippet = "Mr. Crowe gave us a grand tour and then offered detailed advice about
 where we 
  should ... In Killarney, County Kerry, we stayed at the Cahernane Hote
l. ...";
gpage_2 has Snippet = "SHIPPEN TOWNSHIP, bounded by McKean county on the north, Elk county
 on the west, 
  ... 'The phrase first settlers includes Squire Freeman, S
quire Crow, Col. ...";
gpage_3 has Snippet = "(Parish Wandsworth, John Traviss Squire, England Wandsworth (London Jo
hn Traviss 
  ... July 3 Edward Crow & Elixabeth Games, Banns. Sep. 24 John Wright,
 jun., ...";
gpage_4 has Snippet = "Jan 14, 2008 ... Essex County The Corporation of the Town of Ki
ngsville ... Crow, Sarah J. .... 
  Squire, Thomas · Starrett,
 Agnes P. Starrett, James S. ...";
gpage_5 has Snippet = "At this point, the crow lost count and returned to it's nes
t. Not being able to 
  distinguish above the number four cost the crow it's life, as
the squire rid ...";
gpage_6 has Snippet = "Lockhart is the county seat of Caldwell County which is central
ly located just 15 
  ... Dr. Steven Crow of Richardson Pediatrics, and Southwest T
ransplant ...";
gpage_7 has Snippet = "Township, along the southeastern boundary of Chester County,. Pennsylv
ania. 
  The Squire Cheyney parcel is situated within the northeastern portion ..."
;
gpage_8 has Snippet = "Chair Squier inquired about the location of the billboards discussed i
n the loan. 
  request by the Columbia Gorge Discovery Center and Wasco County Museum. 
...";
gpage_9 has Snippet = "2 Squire & Polly Tripp SHAVER, George Henry M Apr. 13 George & Jane Sh
aver ROBBINS, 
  ... 7 William Crow & Susan Lovey MYERS, Catharine Marilla F Jan. <
b>...";
ke$

• Combinations

ke$ . is gpage_5;
ke$ ! toc $URL | wordbag | sort -rn done;
1*special
1*numbers
ke$ ! echo $Snippet | wordbag | sort -rn done;
2*s
2*crow
2*#39
1*this
1*squire
1*rid
1*returned
1*point
1*number
1*nest.
1*lost
1*life
1*four
1*distinguish
1*count
1*cost
1*being
1*at
1*as
1*above
1*able
1*...
ke$