John Grieve: Search engine development    
 Search engine development 15 comments
18 Jun 2010 @ 03:27, by John Grieve

Over the last few years I have had a number of ideas on how to improve existing search engines and develope new ones. I am looking for interested parties to co-operate on implementing these ideas, particularly people with qualifications in Search Engine Software Technology. If you would like to get involved please contact me on

[< Back] [John Grieve]



20 Jun 2010 @ 07:43 by mortimer : drumrobot search engine
While the colloquial term "Semantic Web" is not formally defined, I like to think is a true semantic website; a social bookmarking tool like which allows members to collectively index web pages with tags/terms freely chosen. However, the "search engine" aspect can be improved with some new features about the organization of tags (or interactions between users) and tag relations i.e. WordNet lexical database.

Couple more robot updates want rolled out, but soon the next step is "tag hierarchy" using WordNet.

note; the first incarnate of drum robot used piwik statistics, the memory overhead proved expensive; and so piwik caught the hatchet...same could happen with WordNet  

20 Jun 2010 @ 10:55 by johnjoseph : search engine improvements
Here's a simple way of presenting what I'm saying about search engines. When you search for something very simple on Google you often get a result accompanied by "500,000 results in 0.33 seconds". Obviously there is too much information here. You don't need half a million results to a search "President of the United States". You need to restrict and control and target the search and one obvious way to do this is to specify the location (geographical), and it is noteworthy that ads on Google are already location specific but not the searches themselves. Other ideas suggest themselves. Geddit?  

20 Jun 2010 @ 11:46 by mortimer : Search Operators
yup, goddit.

List of Google Search Operators - {}

location: {}  

20 Jun 2010 @ 12:16 by johnjoseph : search engine improvements#2

If we can combine your detailed technical knowledge of these things with my more general "common sense" approach we may be onto something. I have a lot of ideas much better than the one you answered. Sorry if i was teaching my "granny to suck eggs". john  

23 Jun 2010 @ 11:31 by mortimer : Not at all
I'm more of a hack than "qualified in Search Engine Software Technology". Hack hack stitch together pieces that work etc. The basic search functions are built into PHP and SQL - - the popular choice for web development.

Lot of script kiddies are trying to duplicate the Google search algorithm by using PHP. Which is futile because interpreted language would never work in large environment. For speed and optimization the algorithms should be compiled into native executable.  

23 Jun 2010 @ 11:48 by mortimer : Then fun starts
After you have an executable use the PHP {|exec()} or the {|system()} to return the algorithm.

For example, using exec() function the drum robot thumbnail images are created by an external program -  

23 Jun 2010 @ 11:53 by johnjoseph : bio-feedback
Those electrodes look formidable, can we have some other form of bio-feedback which may point to identifiable wave forms, such as lie-detectors?  

23 Jun 2010 @ 12:03 by johnjoseph : bio-feedback mouse

we could even develop a mouse that functions as a bio-feedback device, as in Lie-detectors, Scientology technology, john  

23 Jun 2010 @ 12:55 by johnjoseph : Non-computability
Would input from a right-brain bio-feedback device , make an algorithm non-computable?  

23 Jun 2010 @ 16:14 by mortimer : Right brain logic
Whether it comes from your left brain, right brain, or comes from your big toe after stubbing it etc. If its practical then its logical.  

23 Jun 2010 @ 16:22 by mortimer : Electrodes look formidable?
Nice graphic for newciv 2.0  

23 Jun 2010 @ 17:48 by johnjoseph : is god's mind a computer

alpha waves relaxation and meditation, their synchronization is a sign of enlightenment,does that compute  

27 Jun 2010 @ 20:17 by johnjoseph : REM 56 specifies agreed location
Following my comment of 20th june and your reply regarding search criteria it occurred to me that instead of using words and grammar to restrict the search it would be easier to employ an easily understood set of (numerical) codes. For example REM 56 could indicate location as in the item you mentioned. The REM preceding the code would tell the search engine that this text was to be ignored and the following number a code as in the REMark instruction in BASIC.Also it would be possible to set default meanings for all relevant codes in a personalized repertoire,so that for example 567 could mean England,56 could mean London and 5678 Europe. Once you had set these default values you could have an override code to mean global/ such as 5.  

27 Jun 2010 @ 20:48 by johnjoseph : exclusion codes/search criteria
More powerfully you could have what i call personalised exclusion codes, so that REM 77 would mean exclude everything connected to right angled triangles or REM 88 would mean exclude anything connected with football. In this way you could easily exclude huge numbers of results and bring the number down to manageable proportions---john  

6 Jul 2016 @ 03:27 by king king @ : king  

Your Name:
Your URL: (or email)
For verification, please type the word you see on the left:

Other entries in
26 Feb 2017 @ 20:56: Trump versus the Media
26 Jan 2017 @ 18:53: Women's Marches the beginning of the Fightback
7 Jan 2017 @ 20:11: Now's the End Times
15 Nov 2016 @ 22:48: Theory of Civilization-- Part 3
15 Nov 2016 @ 22:20: Theory of Civilization-- Part 2
15 Nov 2016 @ 21:52: Theory of Civilization-- Part 1
14 Nov 2016 @ 22:11: Global march of the right in the context of the dynamics of civilization
12 May 2016 @ 21:47: SuperCivilization, The Second Axial Age and not-so-enlightened Despots Cont.
12 May 2016 @ 21:20: SuperCivilization, The Second Axial Age and not-so-enlightened Despots
28 Feb 2016 @ 18:55: Economic Evolution

[< Back] [John Grieve] [PermaLink]?