The explosion of using all forms of social networks, such as Facebook, Twitter, Flickr, leads to an unprecedented content production from which valuable information can be extracted. Specifically, the analysis of users’ emotions as these are imprinted in such social networks can provide information for various parties, such as stakeholders, companies, sociologists or even for the users themselves. The recognition and analysis of such emotional information is known as sentiment/affective analysis, which is the process of detecting sentiments from texts. In other words, it determines the emotions being expressed in a piece of text. In the following sections we describe two sentiment analysis approaches that have been implemented and applied in content which has been extracted from users’ activity in Facebook, the very popular and widely used social network. Within the REVITAL project was conducted a 15-days experiment where the Facebook accounts of 38 users were monitored for retrieving the messages/posts that they have exchanged with their online social connections. With an initial data analysis of such data we can conclude that the most dominant emotions tend to be the emotions of surprise and joy. In this deliverable we provide a comprehensive description of the developed approaches and some initial results that are related to the introduction of such methodologies in content produced by users’ activity in social network.Real life affective experience report
People utilize their social networks constantly on a daily basis in real, face to face, or in virtual social interactions, over the Internet. In such real and virtual social interactions emotions have an important role. People are driven by their emotions when they express opinions, interact and communicate for multiple everyday life tasks. However, the study of naturally occurring emotions in virtual interactions is limited. The REVITAL study adopted a combined methodology to study emotion experience in real and virtual social interactions using: (a) an experience sampling methodology (ESM) of emotion states in real and virtual social interactions, and (b), a lexicon-based approach to identify the emotions expressed in naturally occurring computer-mediated communications and to verify the results extracted from the previously conducted analysis (i.e., offline social activities’ analysis). The focus of this deliverable is the application of the ESM methodology in the RETUNE's data and its respective results.Prototype for web & mobile applications
The REVITAL project aims at contributing in understanding how people’s emotions are formulated in their online and offline social activities. The analysis of people’s emotions in their online social activity can be conducted by analyzing the content provided by them in their social networks. But for the offline social activity the people themselves have to explicitly provide information about how they feel throughout their everyday interactions with their surrounding environment (e.g. family, friends, and co-workers). So, in order to provide users with an easy to use way to manually declare such emotions, we developed a mobile and web application. In this deliverable we describe these applications in relation to their technical and functional characteristics.