Zetta Cloud (www.zettacloud.ro), a company specialized in intelligent data analytics services and digital news software, published a benchmark performed on a large set of online news articles with the purpose of unveiling trends, patterns and characteristics of various online publishers in the context of automated news verification using Artificial Intelligence algorithms.
The benchmark was performed on a set of almost 17,000 web pages using TrustServista (www.trustservista.com) a unique software platform that can automatically determine the origin and trustworthiness of online news.
The analysis is focused on understanding the feasibility of using a fully automated approach to news analysis and verification. The data set was analyzed based on the following metrics:
- Authorship. Which type of news sources declare named authors and which do not.
- Context Setting. How well is the factual information represented in the different types of publications.
- Polarity (Sentiment Analysis). What are the main polarities (negative, neutral, positive) of the content and what is the impact in terms of topics, writeprint and emotional triggering.
- TrustLevel and Clickbait score analysis.
- Patient Zero analysis.