The much awaited software platform that can determine the origin and trustworthiness of online news – TrustServista – will be in Alpha version in just a couple of days! This means that TrustServista will be running in production, with real data. And – more important- we will start allowing selected beta users to use the platform and investigate processed news websites.
The next phase – Beta – will most likely be mid-February and will consist of:
- More news websites added to the platform
- Better data visualization and reporting capabilities
- Improved TrustLevel scoring, based on the beta users feedback.
In the meantime, we are proud to present the first public screenshots of TrustServista and brief explanations on what the platform does.
The main application dashboard allows the verification of a web URL (“Check Article by URL”) or navigating through the most recent trending topics, consisting or similar articles clustered together with the use of artificial intelligence. This is the page where the investigation starts:
Clicking on one of the trending topics will open up the detailed topic dashboard. This dashboard displays the time distribution of news articles from the topic, the distribution of sentiment, the distribution of TrustLevel, the newest and oldest article and the two most important Topic Relationships and Topic Timeline graphs.
The Topic Relationships and Topic Timeline graphs present an overview of all the articles from the current topic, including the links between them, and displayed on a easy to digest timeline:
This dashboard is mainly used to discover trending topics and get an overview of the topic’s characteristics an relationships between its articles.
Selecting any article from the cards or graphs, or checking an article by URL from the previous dashboard will open the Article Dashboard. This screen provides detailed information about a single article, such as:
- An Article Details card, containing an automatic generated summary
- A list of Top Entities extracted automatically from the article: people, locations, dates, institution
- The article’s Patient Zero, which is the article where the information originated:
- The TrustLevel which determines the trustworthiness of the article:
A Patient Zero to Article timeline will display the distance in time between the current article and Patient Zero:
Also, the links between the current article and other articles from the same topic, including Twitter messages, can be analyzed in the Article Graph card. The links displayed are from the 1st to the 3rd level, showing how articles reference other articles, either explicitly (URL reference and link) or implicitly (no link, but textually mentioning other sources).