With the goal of understanding how major public interest events are perceived by the TV public and how the users' interests evolve in time, we introduce a data collection, integration and analysis framework that allows to compute, characterize and explore dynamic social web communities. We focus on communities of Twitter users that interact each-other around specific T V events and track their public cybersocial activities to analyze and visualize the information difusion processes and understand how they afect the community structure of the social network.
People on the Web talk about television. Trending topics and concepts arise from TV users’ “cybersocial” activities thus influencing the community structure of the social network. Users aggregate themselves around topics and concepts that emerge from major events. The U.S. presidential primaries have a clear effect on the community structure of users around the world, but such groups are subject to changes as the candidates pronounce some speech in favor or against some policies (e.g., on immigration, protectionism, welfare) that concern them, directly or indirectly. Capturing and understanding the evolution of such communities promptly is of great value for a number of stockholders: advertisers, broadcast programmers, market and society analysts.
Within the RAI research project on the study of the integration
between social networks and the TV world, the collaboration with
the University of Turin led in the last years to the design and
development of a general framework1 for the collection and analysis of
heterogeneous data coming from standard TV sources (EPGs,
editorial metadata, audience metrics) and social networks conversations
[
Our application enables both guided and automatic extraction
of emergent topics and concepts from Twitter public conversations.
Based on such concepts, it builds the social network of users by
leveraging their interactions. Two users are considered as
interacting each other if they mention each-other explicitly (in a retweet or
a reply) or whenever they refer to similar concepts in their
conversations. Moreover, our application enables the execution of multiple
community detection algorithms [
In the last three years, a number of tools have been proposed to
address the problem of complexity and dynamicity in network
analysis and visualization (eg., [
Our framework consists of three internal layers, covering all the phases from data collection, representation and integration to data analysis, and a visualization layer (see Figure 1). More specifically, a Source processing layer contains the diferent modules for collecting all the data from web sources to be conveyed in the social graph and other data representation technologies. It accesses a number of predefined web/social/media sources (Twitter, oficial web sites, TV channels, ontological information sources) and continuously processes the information collected in real-time to detect the named entities (people, places and events) trough the use of a Named-Entity Recognition module and a topic extraction module. The collected users, topics, concepts and relationships among them are then stored in the social graph layer based on Neo4j2 and MySQL3, which contains all the modules needed to store and manage the social graph. The social query and analysis layer ofers functionalities for querying, browsing and analyzing the graph. More specifically the analysis module provides a set of community detection and social network analysis components.
Community detection if performed by adopting two well-known
algorithms: Louvain [
The visualization layer is constituted of the web application, that adopt modern web technologies to enable responsiveness and highlevel interaction patterns: node.js as web page server, a REST API for database querying, D3.js as graph visualization and browsing library. To cope with the potential huge amount of data, we adopt a caching mechanism on the layout, nodes, links and communities. 4
In our use case scenario, we gathered Twitter data from February 9th to 13th 2016. The data are related to the five episodes of the 2016 edition of the Sanremo Music Festival, the most popular Italian song contest and awards. Overall, more than 2.52M tweets of 176, 760 users have been processed. In these conversations, the relevant concepts are mentioned more than 1.80M times, while the overall number of hashtags is 3.60M. Our goal is to recognize and characterize those communities that gravitates around some specific events that happened during this major national event. In particular, during the 2016 edition, apart from the song contest itself, several events attracted the public interest and were extensively discussed in newspapers, news broadcasting, blogs and social media. For instance, in all episodes, some artists and guests promoted the upcoming law on civil rights for the LGBT community. During the second episode, there was a touching exhibition by Ezio Bosso, an Italian composer known worldwide afected by an autoimmune neurological disorder. In the final episode, Elio e le Storie Tese, a well-known comedy rock band, performed their song dressed as Kiss, the famous metal band; this exhibition was appreciated by Kiss frontman Gene Simmons, thus gaining visibility at international level
A user has the possibility of selecting a set of concepts of interest (e.g., the artists exhibiting during the event, “LGBT”, “Ezio Bosso”, “Kiss”) or a set of “trending/emerging concepts” according to some basic statistics. Then, a community detection algorithm is executed on the Twitter interaction network of the users that mentioned at least one of the selected concepts in their conversations (also including replies and retweets). The network and their communities are visualized as shown in Figure 2. Each community can be then inspected individually and the application returns some related statistics: the list of most influential users (according to pagerank, betweenness centrality or other network metrics) the list of the most relevant concepts and hashtags (identified by a summarization technique) including both selected concepts and other co-occurrent concepts. In Fig. 2, for instance, Community 2 is characterized by the fact that the involved users talk about Dear Jack (an Italian rock band), while the hashtag #MezzoRespiro refers to the title of the song presented during the contest. Community 5, instead, is a broad interest community dealing with Ezio Bosso and the civil rights of LGBT people (#SanremoArcobaleno).
The application supports also the analysis of the traces of the information difusion processes involving individual users and communities (see Figure 3). By retracing the timeline of each episode, the application allows the user to visualize and compare the spread of a number of selected concepts in the interaction network. During each time interval, Twitter users that tweet, retweet or reply in conversations mentioning a specific concept c are “turned on” and highlighted with the color identifying c. In this visualization modality, the information difusion process can be compared with the overall Twitter activity represented by the curve on bottom of Figure 3. This facility supports the discovery of difusion patterns, as well as the visual comparison of difusion trends. Notice that basic complex network statistics (number of nodes/edges, number of connected components, clustering coeficient) are always available on screen. Moreover, the user can observe the time evolution of such metrics referring only to the highlighted subgraph. Though usable on mobile devices, the application is optimized for 4K UHD displays.
Through the described platform, it is possible to identify and visualize phenomena that occur on social networks with reference to the television world such as emerging phenomena, their temporal evolution and the relationships between the entities involved. The conceptual navigation network allows the user to walk through the concepts and involved social users mentioned within television events in a particular timeline based on mentions of mentions during television events.
The flow of data coming from social networks is now seen as a true additional information channel for the end user, for some television programs it is a fruition line for some verse independent of the television program it is generated. The platform described, in addition to being used as a support to marketing choices, can be used as a data source to tell television events from the point of view of users who have lived and commented on them, creating an unseen story that sees as protagonists the viewers themselves with their perception and emotions tried during the viewing of a television program. In fact, Digital Storytelling, or Narrative made with digital instruments, consists of organizing selected content from the web in a coherent system based on a narrative structure, in order to obtain a tale composed of multiple elements of various formats (video, audio, images, texts, maps, etc.). Digital Storytelling is therefore a form of narrative particularly suitable for communicative forms such as journalism, politics, marketing, autobiography, didactics and it could became a new frontiers of the television world to provide information and entertainment experiences that can be visualized through diferent tools of fruition in synergy with each other. 5
In this paper we presented an application that allows to find, characterize and explore community of users in social networks conversations. We demonstrate the efectiveness of the approach by exploiting a use case in the TV setting related to a well known Italian song festival. The collaboration between University of Turin and RAI, the Italian broadcaster, is constantly evolving: in the next months, we will integrate other sources that we are already monitoring (e.g., Facebook), as well as new features for a more in-depth characterization and comparison of users and communities.