Social media

Two systems will be developed for the application on social media of WP3: a web interface to display the analysis of the public discourse (i.e. a collection of news items); a plug-in to contextualise news-related posts within the public discourse.

Main aim

The web application aims at exploiting textual features (style, semantic, moral) to group, visualise, summarise, and explore news.

Users

The application is targeted to news professionals and monitoring authorities.

Description

The user can filter the corpus of news items (we will provide) by time period, topic and presence of keywords. On these selected documents, textual features are extracted and used to organise/cluster such data. This analysis is then visually returned to the user through a web interface. The aim is to show the main message and structure of the public discourse (on selected time and topic). In particular, the user will be able to inspect documents (or salient ones) freely or following exploration paths based on similarity, dissimilarity (of moral values, style, cluster, etc.). Based on these functionalities, news professionals can explore the news, selecting documents specific topics and writing review reports on them. Moreover, authorities can parse topics looking for hallmarks of polarisation, e.g. presence of clusters with contrasting moral values.

User input

Corpus' time interval, presence of keywords, topic, interaction with plots and recommended news items.

Output

Visual representation of corpus’s documents by clusters; analysis of clusters (summary); recommendation path from selected document, and from distant (from other clusters) documents