Filter accurate and potentially valuable information from the social noise

Promptly and accurately discover, analyze, and verify the credibility and truthfulness of reported events, news and multimedia content that emerge in social media in real time.

Introduction to TruthNest

The growth of social media and the dominance of smart mobile devices are currently reshaping the nature of journalism. On the reporting side, citizen journalism flourishes. On the consumption side, a large fraction of conventional media users turn to social media for information purposes, due to their direct nature and the collaborative filtering that takes place in their social circles.

These two sides reinforce the use of social media for informational purposes. However, they also fuel the spread of rumors and false news. In the “word-of-mouth 2.0” age it is particularly difficult to distinguish what’s accurate amongst the noise. During the short period just after the emergence of a news story in social media and before it actually spreads, it’s extremely critical to validate the authenticity/truthfulness of the story. This is the gap TruthNest has tried to fill.

TruthNest employs the latest technology to deliver a holistic analysis including a plethora of social dimensions, automatic controls and metrics in order to enable media organizations and journalists to:

  • Identify early signals of events in the critical interval between the first minutes after an event occurs and before the social media channels are flooded with noise.
  • Verify authenticity of information posted, a particularly challenging and time-consuming process during turmoil situations where information flow is blocked between controlled media and social media polarization.
  • Locate possible credible sources of information efficiently.
  • Monitor the social ecosystem effectively by creating smart, semantically meaningful, context-aware, dynamic, cross-network streams.


Detect emerging stories early in social media, before they hit the mainstream news


Reveal a wealth of properties about the source and the content with one-click, using TruthNest deep comprehensive analysis


Use TruthNest’s credibility score to rate emerging information, based on metrics regarding the contributor, content and context

Identify key events and stories in the social noise

  • Create custom streams of information to stay tuned to your subjects of interest.
  • Search across channels with a variety of filters including location, time, user profile or other quantitative information.
  • Discover relevance and structure in the social information storm with named entities extraction and post categorization. See the technology insight section for more information.
  • Monitor the social ecosystem effectively by creating smart, semantically meaningful, context-aware, dynamic streams
  • Manage your streams seamlessly in one place, including diverse social items from timelines, public lists, mentions, hash tags, etc.

Trigger a holistic comprehensive analysis, delivered in seconds

  • Take advantage of TruthNest’s "3C-analysis" (3C stands for the triplet: Contributor, Content and Context) which quantifies a wide array of characteristics of the source, the multimedia content and the context for each post.
  • Use the 3C-analysis to quantify the reputation, history, popularity, influence and many more social dimensions for a complete understanding of a source’s position in the community and a complete profile of their activity.
  • Exploit semantic technologies such as sentiment analysis, named entity recognition and event identification for meaningful, context-aware results. For more details, see our technology section or feel free to contact us.
  • Make the most of the 3C-analysis results through a consistent, human-friendly system of indices, designed for the non-specialist. Exploit our rich interactive visualizations that aim to maximize your insight.

Verify posts including their multimedia content

  • TruthNest offers support for your verification process via:
    • Semantic metrics, which derive the context of each post (e.g. persons, locations and organizations mentioned, the topics it refers to, the collective sentiment around it etc.)
    • Network quantitative metrics, indicative of the status of a source in the community (e.g. its influence) or the penetration of a post in a community (e.g. trending topics).
    • Multimedia forensics metrics, which reveal the attached media authenticity and originality.
  • Benefit from the TruthNest’s truth score. For the analysis of each framework category, we define a set of related parameters: reputation, history, popularity, influence, originality, authenticity, etc. Each modality is rated independently and individual scores are combined to provide an aggregate truth score.

TruthNest’s state-of-the-art technology

TruthNest uses state-of-the-art Machine Learning and Quantitative Analysis technologies to deliver valuable insight concerning both semantics and network-level properties that matter to the media professional. We introduce several innovations, specifically designed to improve journalistic work.

TruthNest is powered by an Artificial Intelligence engine that handles Natural Language Processing and a Quantitative engine that crunches a massive flow of Big Data from social media.

  • The Artificial Intelligence engine enables semantic synthesis:
    • Named Entity Recognition identifies entities such as persons, organisations and places within posts as well as certain semantic attributes that describe them.
    • Classification categorizes posts into broad categories (e.g. politics, sports, business, technology etc.) using supervised machine learning techniques such as Support Vector Machines and Neural Networks.
    • Sentiment analysis uses classification (into positive, negative and neutral) together with sentiment lexicons in order to compute how positively/negatively a post is perceived within the community. The strength of the sentiment of the responses to a particular post may be indicative of whether the post is credible or interesting for further investigation.
    • Topic extraction and event identification uses unsupervised machine learning algorithms to cluster posts in order to compute those metrics that demand topic and event awareness. Topics are narrower than the broad categories computed with classification, and refer to more specific subjects.

About ATC

Social Media analytics technology involves a great level of innovation and applied science. ATC has an extensive experience in large scale R&D projects: We have led and directed several European Commission projects and partnered with international Industry and Media leaders as well as top-tier universities. As a consequence, we possess a deep knowledge and understanding of innovative technologies and the Science that propels them, often as early as from the very point that they are born. We contribute to the vision and the evolution of these technologies, instead of merely being early adopters.


For more information or for arranging a demonstration of the service please send us a message

Thank you! Your submission has been received!

Oops! Something went wrong while submitting the form :(

Visit Us

Rizariou 10
Chalandri 152 33, Athens

Mail us

Call Us

T: +30 210 68 74 300