Thursday, September 25, 2008

Visualisation Complete

I have used the visualisation tool that is included within the carrot2 project.

It is basically a flash visualisation that will display the main clusters as a pie graph and also display the results in that cluster in the frame next to the flash object.

Here's a snapshot of it in action, the search query was 'john howard' and that was run through

Monday, September 22, 2008

Intergration of the VisionBytes database search complete

I have integrated my app into the carrot2 webapp. My app simply creates a connection to a mySQL database and uses;

String sqlQuery = "SELECT * FROM segment WHERE summary LIKE '%" + query + "%'";

as the main search query.

It the will create a list of Documents with the appropriate fields populated from the mySQL query and that list will then be passed onto the next component of the application (the clustering engine).

Thursday, September 11, 2008

mySQL troubles

After what took like ages converting the VisionBytes Data (in the form of SQL statements) into mySQL statements, and keeping all the relationships intact, I wrote a simple app to compare and retrieve the appropriate data from the database from a query (a user inputted string).

Basically, the table that is of interest is the 'segment' table. The reason for this is because it contains;


Both of which is used to determine the location of a story/topic


Both of these are necessary to create a 'Document' in the webapp of the carrot2 project.

The breakdown of a 'Document' is as follows:

Document(String title, String summary, String contentUrl)

In our case, we will replace url data with the location of the topic/story.

Friday, September 5, 2008

Draft treatise handed in

I was almost late handing the draft in today, luckily I allocated enough time for travel.

What normally takes about 20mins to reach uni took about 40mins due to traffic.

You know what they say, it pays to plan ahead :)

Thursday, September 4, 2008

Draft treatise almost completed

Basically the chapters that I've worked on for this draft is:

-Problem Statement
-Aim and Scope
-Thesis Overview

-History of Topic Detection and Tracking
-Previous Thesis
-Approaches to Integrating the Search Query Component
-Approaches to Integrating the Clustering Algorithms/Engine
-Approaches to Integrating the Visualisation Component

-Search Query Component

-Conclusion and Possible Future Work

But there's still a lot more work to be done...

Just another small update:

-Google search is no longer supported by the Carrot2 project, however the Yahoo web search, MSN live, as well as is supported.
-The Yahoo web search is already integrated in the carrot2 project