CReMA - Customer RElationship Management Application - is a proof of concept about social network analysis written in JavaScript.

The initial page of CReMA display an input field filled with the url of a forum, the "Getting Started with Studio" section of Aptana forum. You can scrape any forum based on phpBB forum solution: the url is not the root of the forum, but it is referred to a section, containing the topics list.

Pressing the "Search" button, the forum is scraped, retrieving authors, coauthors and tags information.

The scraping phase collects data using only the first page of the section - about 40 posts without sticky posts - and only the first page of each post. The process should take less than one minute and the display is updated for each post. The result is cached client-side expiring every day.

After the scraping phase, CReMA displays two tag clouds and two social networks graphs.

The left cloud contains the authors and the coauthors of the posts. The right one contains the tags of the posts. The left graph displays the relationship between the post author and his coauthors. The right one displays the relationship between the post author and the content of the post, summarized with a tag.

You can hover the mouse on the graph nodes, displaying the name of the author, coauthor or tags. Clicking on a node of the author graph, you can higlight the tags related with the author; or clicking on a node of the tag graph, you can highlight the author of the post. Clicking on the cloud the behaviour is similar. Furthermore if you double click on a node, you can go to the post related with the author or the tag.

Author, coauthor and content of the post is scraped using jQuery server-side. The content of the post is tagged with the Term Extraction Documentation for Yahoo! Search service. The layout of the graphs is calculated implementing the Fruchterman-Reingold algorithm.