From Media Noise to Meaning: NewsVibe’s Research Goes to GOR 2025

We at NewsVibe are proud to announce an important milestone in the development of our platform: a key component of our innovative network analysis method, at the core of one of our upcoming features, has been successfully presented at the prestigious General Online Research (GOR 25) academic conference, held between March 31 and April 2, 2025, at Freie Universität Berlin. This academic milestone reinforces our commitment to combining advanced research with practical media intelligence tools.

The research, titled “Where is Everybody? Measuring Semantic Source Position and Creating Online Discourse Typologies from Co-Occurrence Networks,” was developed by Dan Sultanescu, Pavel Dimitrov Chachev, and Dana C. Sultanescu from Social Monitor, the expert team behind NewsVibe.

At its core, our method leverages advanced network analysis techniques to map how online media sources contribute to and shape public discourse. Using real-world data extracted via the NewsVibe platform – specifically 30 topics relevant for the US media environment, extracted from content published by relevant US online sources – we built entity co-occurrence networks, identifying and quantifying relationships among the most relevant entities (people, organizations, places) in online news and social media conversations.

The main stages of the analysis:

1. Data collection via precise queries:
   Using NewsVibe’s powerful engine, researchers extracted content across carefully curated sources, including web news articles and Facebook posts, focused around specific queries (e.g., “Elon Musk”, “JD Vance”, “climate” or “Super Bowl”).
2. Building semantic networks:
   NewsVibe’s advanced AI and machine-learning models first extracted entities from the monitored data. Then we constructed detailed semantic networks showing how entities co-occur in online media narratives, representing distinct thematic clusters or discourse topics.
3. Identifying source positions:
   Applying Otsu’s algorithm to the distribution of source shares across thematic clusters, the analysis identified significant source-topic combinations, clearly distinguishing sources by their unique thematic focus or breadth.
4. Quantifying source diversity with entropy:
   The method computes the entropy of source-topic combinations to quantify the diversity and specialization of news sources, clearly differentiating between highly specialized, moderate, and diverse coverage patterns. For example:
• The network based on content related to Elon Musk had low entropy, indicating strong thematic specialization around a few consistent narratives.
• The JD Vance network showed moderate entropy, reflecting distinct but overlapping political and international narratives.
• The network based on content mentioning Facebook exhibited high entropy, demonstrating a more diverse range of narratives and thematic approaches across news outlets.

• Expertly curated sources:
  NewsVibe selects only premium online and social media sources from Romania, the US, Germany, and Moldova, curated by experts with over 20 years of experience in media and communication. This ensures relevance and quality, free from the typical online noise.

• Real-time, mobile-friendly access:
  Users have immediate access to insights wherever they are—without needing specialized analytical teams. NewsVibe translates complex data into clear, actionable intelligence instantly accessible on any device.

• Cutting-edge AI and proprietary algorithms:
  The platform’s proprietary Machine Learning algorithms perform advanced tasks including filtering, content classification, Named Entity Recognition (NER), advanced sentiment analysis and sophisticated cluster analysis (Smart Topics).

• Dynamic analytics and strategic insights:
  NewsVibe offers dynamic analytics dashboards showing real-time trends, impact analysis, key influencers, and deep sentiment insights (through sentiment analysis, hashtags, and emoji analysis) – helping users understand not just the content, but the emotional tone and narrative direction of media conversations.

• Strategic decision support:
  The platform aids strategic decision-making in crisis response, brand and reputation management, competitive benchmarking, and comprehensive understanding of media coverage. It provides precise, actionable insights translated into strategic intelligence.

The method presented at GOR 2025 directly informs a powerful new capability soon to be integrated into the NewsVibe platform: advanced network analysis of media discourse. Users will be able to visually explore semantic networks around topics of interest, clearly mapping influential sources, thematic clusters, polarization points, and narrative dynamics.

Imagine entering a query about a critical event or public figure and immediately seeing a dynamic visualization showing the clustering of entities within the dialogue or how various sources position themselves around different thematic clusters – this is the near future for NewsVibe users.