A background in psychology would be great but is not essential. Statistical knowledge is essential however.
The App should collect text from the users emails and do content analysis on the text in the messages. It is then graphically presented on an external site and can be compared to all friends the user had contact with. There should be analysis of the text for each person the user had mail contact with (probably only if sufficient text has been written, say 500 words)
also, a similar analysis is done for text on the wall, again splitting own wall posts and that of friends. The user is then forwarded to an external page where he can compare his score to that of their friends in detail
this will result in Top lists of friends that you have the most traits in common etc. and those lists can be posted on the wall if the user wants to.
similar to [url removed, login to view] (see graphic), the text in the facebook messages will be used to derive traits of the user and then compare them between his friends. (so imagine a second bar for each trait in the picture to represent the person you want to compare your data with)
All facebook texts will be analyzed and then compared. Using the Regressive Imagery Dictionary [url removed, login to view] and Linguistic Inquiry and Word Count (LIWC) [url removed, login to view] for content analysis.
analyzed data (frequency of certain words etc.) will be stored in a database (Mysql or similar), not the text!
Users can access an external page at any time for more detailed analysis. And to get top 10 lists of friends he is most similar with etc.
Developer is asked to provide for the described functionality and presentation of the statistics seen in the 750 image. We envision this to be an ongoing project as statistical analysis will provide us with new and interesting correlations.
Daily operation of the service should be completely automated.
12 years of perl MYSQL and web. Have some similar project done. Please contact me, for details. Also have pretty basic knowledge in psychology. Interesting project, I'd like to work on it