Automated Analysis of e-Participation Data by Utilizing Associative Networks, Spreading Activation and Unsupervised Learning

According to [1], the term e-participation is defined as "the use of information and communication technologies to broaden and deepen political participation by enabling citizens to connect with one another and with their elected representatives". This definition sounds quite simple and logical, but when considering the implementation of such a service in a real world scenario, it is obvious that it is not possible to evaluate messages, which are generated by thousands of citizens, by hand. Such documents need to be read and analyzed by experts with the required in-depth domain knowledge. In order to enable this analysis process and thereby to increase the number of possible e-particpation applications, we need to provide these experts with automated analysis tools that cluster, pre-screen and pre-evaluate public opinions and public contributions. In this paper we present a framework based on Machine Learning-(ML) and Artificial Intelligence-(AI) techniques that are capable of various analysis mechanisms such as unsupervised clustering of yet unread documents, searching for related concepts within documents and the description of relations between terms. To finish, we show how the proposed framework can be applied to real world data taken from the Austrian e-participation platform mitmachen.at.