Skip to content

Normalized Google Distance for other search engines, like Wikipedia and HN search.

Notifications You must be signed in to change notification settings

LMM0709/normalized-searchengine-index-distance

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Normalized Searchengine Index Distance

Normalized Searchengine Index Distance (NSID), also called Normalized Web Distance (NWD), is a simple generalization of Normalized Google Distance based on the paper Automatic meaning discovery using Google [pdf]. Normalized Google Distance is derived from Normalized Compression Distance, which is derived from Normalized Information Distance, which in turn is derived from the principles of Kolmogorov Complexity.

Machine learning techniques like random forests and SVM's are used to classify, cluster, auto-translate and more.

Searchengine Indexes supported at the moment are Hackernews search and Wikipedia search, but any sufficiently large searchengine index would suffice, provided it returns the number of results found for terms.

The collection of files here is to research NSID globally, and specifically how specialized searchengine indexes influence distances.

##randomforestclassifier.py## ###Description### Predicts semantic relationships between a term and a trainingset, storing them in first-order logic. Currently trained to predict semantic relation to "colors", fed with data from the Hackernews Search API (HN Search). ###Dependancies### urllib2, json, math, numpy, sklearn. ###Sample output### semantic_relation(crimson,colors)[0.23] semantic_relation(rose,colors)[0.21] semantic_relation(tangerine,colors)[0.41] semantic_relation(grey,colors)[0.78] semantic_relation(pastel,colors)[0.48] semantic_relation(design,colors)[0.08] semantic_relation(colors,colors)[0.83] semantic_relation(white,colors)[0.8] semantic_relation(contrast,colors)[0.5] semantic_relation(beige,colors)[0.32] semantic_relation(lilac,colors)[0.21]

About

Normalized Google Distance for other search engines, like Wikipedia and HN search.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%