In the age of Web 2.0 as users actively communicate, interact, and share multimedia content on the web, it become a very interesting research field the use of web mining techniques for discovering patterns in social media sites. Web mining refers to the use of data mining techniques on web data. Although data mining is a well established field, the application of data mining techniques on web data is not an easy convertible task as web data, unlikely the well described and organized data in various databases are usually semi-structured data. A web mining approach in social media sites is to use their APIs for extracting web data from a social media site in order to create data warehouses, where data mining techniques may be applied. This study is build upon a previous work where we analyzed the use of Facebook by Greek parties, where harvesting Facebook data was a laborious task mainly because it was carried out manually. This drove us to the need of using computer programs for automating the collecting data process if we want to constantly monitor parties’ activities on their Facebook Walls. This work will concentrate on working with Facebook’s API for creating a data warehouse consisting of Greek parties Facebook data. This would enable a more efficient analysis, which will extend findings from our previous work. We could for instance analyze citizen activities on party walls determining the users who engage in posting in more than one party wall, identify whether there exist citizen campaigning or the majority of posts are derived from a small number of users posting on a constant basis, identify citizen activities according to their gender and their provided profiles etc. Creating such data warehouse will also enable us to use data mining for discovering interesting political communication patterns.
