Topic Tomographies (TopTom): a visual approach to distill information from media streams

Abstract

In this paper we present Top Tom, a digital platform whose goal is to provide analytical and visual solutions for the exploration of a dynamic corpus of user‐generated messages and media articles, with the aim of i) distilling the information from thousands of documents in a low‐dimensional space of explainable topics, ii) cluster them in a hierarchical fashion while allowing to drill down to details and stories as constituents of the topics, iii) spotting trends and anomalies. Top Tom implements a batch processing pipeline able to run both in near‐real time with time stamped data from streaming sources and on historical data with a temporal dimension in a cold start mode. The resulting output unfolds along three main axes: time, volume and semantic similarity (i.e. topic hierarchical aggregation). To allow the browsing of data in a multiscale fashion and the identification of anomalous behaviors, three visual metaphors were adopted from biological and medical fields to design visualizations, i.e. the flowing of particles in a coherent stream, tomographic cross sectioning and contrast‐like analysis of biological tissues. The platform interface is composed by three main visualizations with coherent and smooth navigation interactions: calendar view, flow view, and temporal cut view. The integration of these three visual models with the multiscale analytic pipeline proposes a novel system for the identification and exploration of topics from unstructured texts. We evaluated the system using a collection of documents about the emerging opioid epidemics in the United States.

Collective response to the media coverage of COVID-19 Pandemic on Reddit and Wikipedia

“…Although there are many approaches to temporal and hierarchical topic modeling [104][105][106], we choose to apply NMF to the dataset, and then build time-varying intensities for each topic using the articles publication date. Starting from a dataset D containing the news articles shared in Reddit, we extract words and phrases with the methodology described in [107], discarding terms with frequency below 10, to form a vocabulary V with around 60k terms.…”

Collective Response to Media Coverage of the COVID-19 Pandemic on Reddit and Wikipedia: Mixed-Methods Analysis (Preprint)

“…Nonnegative matrix factorization (NMF) [67] is the counterpart of LDA for matrix factorization. Although there are many approaches to temporal and hierarchical topic modeling [68][69][70], we chose to apply NMF to the data set and then build time-varying intensities for each topic using the publication dates of the articles. Starting from a data set D containing the news articles shared in Reddit, we extracted words and phrases with the methodology described in [71], discarding terms with frequencies >10, to form a vocabulary V with approximately 60,000 terms.…”