Hercule: Journalism Platform For Finding and Fact-checking Breaking News

In today’s digital world, everyone is connected and has access to huge amounts of information. News travel fast and bad or fake news travel fastest.

People in social media frequently share or retweet information they have not checked thoroughly. As a result, sensational news stories often take the spotlight, grabbing readers’ attention.

A Wild Goose Chase

In this environment and in our increasingly divisive world, journalists are overwhelmed with conflicting, contradictory rumors that may or may not turn out to be a topic worth covering. Twitter is the go-to social media to check for what’s trending and Google News – for what various media outlets are reporting on a topic.

But waiting to see what’s trending on Twitter, looking for hashtags or keeping an eye on Google News and YouTube is time and resource consuming. In the fast-paced media world today, if journalists want to keep up with the competition, they need to search and discover emerging hot topics quickly. They also need to add value to the breaking news by enriching them with additional points of view, sentiments and background information.

Tracking Down Breaking News

That’s why we at Ontotext have been developing Hercule: a platform to help journalists detect emerging news topics, check their veracity, track an event and find the various angles in a story as it unfolds.

 Hercule Index Page

The platform is the result of the PHEME project, which was funded under the European Union’s Seventh Framework Programme for research, technological development and demonstration. Its goal has been to create a computational framework for automatic discovery and verification of information at scale and fast. This framework would enable a better understanding of the impact rumors have in social media: where they originate, how they spread, who and why spreads them, with multiple uses in many domains.

Compared to searching Twitter, YouTube and Google News, the Hercule platform has no limit to the number of results journalists can review – it is able to process big amounts of Twitter data as well as updates from YouTube and Google News. In addition, the platform automatically detects the meaning of the sources and links, which have been shared, and clusters them in storylines. That’s because of the semantic technology we use in our products and services. The platform users can easily scan the automatically clustered storylines and detect emerging stories. They can also check the checkworthiness of the information that has been shared based on a predefined score.

The Road Ahead

The Hercule platform interface is still being finalized but we have planned and developed many exciting functionalities.

For example, the UI would be able to identify stories and allow journalists to create their own topics by entering keywords they want to be tracked in tweets. Once the topic is set up, Hercule automatically creates and clusters the results in the so-called “storylines”, enabling journalists to review the most popular topics created, the recently viewed topics or the emerging topics.

The storylines also can help journalists to identify emerging hot topics. If someone stumbles upon an unfamiliar concept such as a name or a place, the “related news” section of the platform links it to NOW – our demonstrator for semantically-enriched publishing. It contains easy to browse and check concepts of people, organizations and locations.

Automatically generated Storylines

Another functionality of the interface focuses more on the investigative aspect of news gathering and dissemination. It helps journalists review how larger topics have unfolded. They can see the timeline of the Twitter activity and review the popularity, veracity and checkworthiness of the topic they have created.

Each of these features have scores, pointing to when the top activity on a given topic was, is the Twitter rumor true, or how checkworthy a story is. The checkworthiness score combines two features: how easy it is to check the information with a trustworthy source and how interesting the topic was for people, so that they have tweeted or retweeted various links about it. This helps journalists enrich their news pieces with various sentiments and opinions, thus adding value to their reporting.

All in all, Hercule will help journalists, editors and editors-in-chief with their primary tasks to identify and investigate what’s new, what’s emerging and what’s worth reporting. The platform will allow journalists to see which topics are of the greatest interest to people and therefore worth tracking and delivering.