Descripción del proyecto
Una nueva interfaz para proteger contra las noticias falsas
Solo se necesita que una noticia falsa se haga viral para que la reputación de una empresa, un gobierno o una persona se vea gravemente dañada. La supervisión es un tipo fundamental de defensa contra los miles de fuentes de noticias que publican información falsa. El proyecto Factmata, financiado con fondos europeos, desarrollará algoritmos y una interfaz inteligente para detectar y agrupar afirmaciones, rumores, opiniones, argumentos y posturas en las redes sociales, noticias y blogs. Esto ayudará a los analistas de campañas de relaciones públicas a obtener información sobre cómo la opinión pública está cambiando en discusiones sobre políticas clave. Esto no solo optimizará su estrategia de comunicación, sino que también les protegerá contra los rumores negativos sobre sus clientes que pueden dañar su reputación.
Objetivo
Organisations (governments/brands) face major issues and are finding it increasingly difficult to tackle information warfare and emerging defamation threats on social media. Rumour mongering, false claims, and aggressively toxic opinions online have severe detrimental effects. Many policy and product launches are executed without granular awareness of how they might be received according to what stance the public is taking, and arguments the public is making in relation to a specific product/policy issue. Stock prices can also take a major hit. Many organisations (PR/Financial analysts/Governments/Agencies) use media analysis to monitor mentions of themselves online; however, current social listening and media monitoring tools are limited: they cannot perform sophisticated detection of key arguments and who is making them, and cannot thus identify useful influencers; they offer poor quality sentiment analysis via keyword detection which does not give insight as to the reason for sentiment, nor explain what rumour or claim caused it to change. Factmata aims to solve this challenge through development of an intelligent media analysis system that can detect and cluster claims, rumours, opinions, arguments, and stances in social media, news and blogs using natural language processing. Unlike current solutions which are limited by simple keyword analysis of media to provide tracking around sentiments, our solution provides algorithms and an interface that can detect, cluster and track claims, rumours, opinions, arguments and stances toward key topics (such as medical/clinical topics or political policy related topics). This enables PR campaign analysts to fully understand how public opinion is shifting on key policy discussions and thus optimise PR and social media messaging and also protect against negative rumours about their clients that can damage their reputation, which is a major risk in today’s “fake news” crisis with the rise of bot and troll networks online.
Ámbito científico
Programa(s)
Convocatoria de propuestas
Consulte otros proyectos de esta convocatoriaConvocatoria de subcontratación
H2020-SMEInst-2018-2020-1
Régimen de financiación
SME-1 - SME instrument phase 1Coordinador
E1 7PT LONDON
Reino Unido
Organización definida por ella misma como pequeña y mediana empresa (pyme) en el momento de la firma del acuerdo de subvención.