Visual thesaurus using arbor.js

Wow, I can’t believe we are at the end of January – one twelfth of the year gone already! I’ve been so busy this month I’ve barely had time to sit down and work on any personal projects. Thankfully I had a spare few hours over the weekend to work on a project I’ve been meaning to finish for a couple of weeks: a visual thesaurus using JavaScript.

You may (or may not) know of the Visual Thesaurus by Thinkmap; very useful but it’s a paid for service, I use it so infrequently I can’t justify paying for it (there is a free trial though). Recently a very interesting graph visualisation library popped up on delicious called arbor.js, so I decided to have a closer look and see if I could use it for the JavaScript version. For the word data I’ve used the Bighugelabs thesaurus API which I previously used to create my Post Thesaurus WordPress plug-in (shameless plug!).

Arbor.js took a little getting used to as it uses a mixture jQuery, canvas and its own methods; but there’s a fairly extensive set of documentation and a set of usage examples. Once you get your head around adding nodes and edges to the system, arbor takes care of the rest. One thing I did notice is the site examples and my VT experiment run much smoother in Firefox than they do in Chrome, which is quite surprising considering how quick the V8 engine is in Chrome. Maybe this will be improved in future versions.

So to the demo; I’ve tried to keep it very simple to use. Enter the word you wish to visualise and click search. Once the API returns a result it’s plotted by arbor. You can drill down into the result by clicking on the sub-particles (and remove them by clicking again). The individual words can be clicked and you’re given the option to search again using the selected word. I still haven’t got the perfect system setup as some results can be a little jittery; usually when an even set of words are returned. I’ll keep playing with system settings to see if I can improve this.

Any suggestions, comments or feature improvements please leave a comment below!