Quicklinks: Introduction - Past - Present - Future - Credits

Introduction

Get-Keywords.com hosts the Get-Keywords Engine, a computational linguistic application that is able to extract the most prominent keywords from a given text. It supports a variety of languages and handles plain text documents as well as hypertext and direct input.

A general purpose keyword extraction engine can be quiete handy, for example to optimize websites for search engines ("SEO"), display an article's keywords to its readers or for scientific purposes.


The past, ...

Initially, in April 2006, the Get-Keywords Engine has been implemented by Alex Linke as his B.A. thesis in computational linguistics at the Ruhr-University Bochum.

Back then, the focus was to create a high-performance application that is funded on a strong theoretical basis. The main goal was to combine the strenghts of statistical motivated algorithms as well as symbolical methods of natural language processing.

To test and prove the concept, the Get-Keywords Engine provided all functionality to process various resources, like plain text and markup enriched document formats like HTML, as well as the four of the top five most frequent spoken european languages: German, French, English and Spanish.

Get-Keywords.com at that time provided a webfrontend to all underlying functions and allowed to process text from files, URLs and direct keyboard input as well as necessary information on usage.

Keyboard input   URL input   File input   Extraction Result

Version 1.0, the last version suited for scientific purposes only, was released end of June 2006 and called "The rocky beach".


...the present...

Based on the implementation and the idea to turn the mostly scientifically oriented Get-Keywords Engine to an actually useful application for the public, Alex and Rona Linke started developing applets on top of this performant engine.

As a first result, Elkx-Box, the "embeddable multilingual keyword extraction box", was created. This application empowers webmasters and/or bloggers to embed all relevant keywords of a webpage in a box on the very page.

Beside that, the Get-Keywords Engine is still actively developed and reached version 1.1 called "Beastie" in early February 2007, featuring improved language data, a more stable and secure way of retrieving text from URLs and code cleanups.

...and the future.

We think about providing public access to the XML based Get-Keywords API, so anybody interested could use it in his/her own application where applicable. An example query result could look like this:

Possible XML RPC example

Maybe we will decide to use XML RPC for this purpose.

However, that's all still up in the air...

Credits

Thanks to Ulrike Britscho for her help on integrating French and Spanish language data! Without her support I wouldn't have been able to integrate support for these important languages.