Reuters tracer: Toward automated news production using large scale social media data

Abstract

To deal with the sheer volume of information and gain competitive advantage, the news industry has started to explore and invest in news automation. In this paper, we present Reuters Tracer, a system that automates end-to-end news production using Twitter data. It is capable of detecting, classifying, annotating, and disseminating news in real time for Reuters journalists without manual intervention. In contrast to other similar systems, Tracer is topic and domain agnostic. It has a bottom-up approach to news detection, and does not rely on a predefined set of sources or subjects. Instead, it identifies emerging conversations from 12+ million tweets per day and selects those that are news-like. Then, it contextualizes each story by adding a summary and a topic to it, estimating its newsworthiness, veracity, novelty, and scope, and geotags it. Designing algorithms to generate news that meets the standards of Reuters journalists in accuracy and timeliness is quite challenging. But Tracer is able to achieve competitive precision, recall, timeliness, and veracity on news detection and delivery. In this paper, we reveal our key algorithm designs and evaluations that helped us achieve this goal, and lessons learned along the way.

Search as News Curator

“…Recent research has explored the encoding of journalistic news values in algorithms that support news content production [23,24,25,26,27]. One example related to curation is Park et al [24], which developed a set of algorithmically applied editorial criteria, such as article and conversational relevance, that were useful to news comment moderators in identifying high-quality comments.…”

DSC-IIT ISM at WNUT-2020 Task 2: Detection of COVID-19 informative tweets using RoBERTa

“…Some works proposed use of SVM (Malmasi and Zampieri, 2018), logistic regression (Davidson et al, 2017), random forest classifier (Burnap and Williams, 2015) and word embedding based method (Badjatiya et al, 2017) for classification of tweet contents. Liu et al (2017) provided the use of unsupervised methods to cluster news topics from tweets.…”

Predicting the success of news

“…[40] One of the best-known examples of computational tools assisting newsroom in social media monitoring and also predicting the newsworthiness of the information is Reuters Tracer, whose development has been openly documented in academic papers. [50][51][52] Topic and event detection are another relevant research areas of predictive analytics in the field of news media. With predictive analytics looking the topic or event as a feature, it might be helpful for the newsroom both to monitor the environment and find newsworthy events and monitor the behaviour of the audience and predict their interests, perhaps even matching these two approaches.…”