Fake news in the cross-hairs of EIT Digital
EIT Digital Social SenseMaking service assesses causes and effects of web chatter.
The EIT Digital SenseMaking Service Innovation Activity works for launching a Social SenseMaking service capable of determining the causes and effects of web chatter, such as Twitter, Tumblr and blogs.
The service delivers an analysis by making text data quantitatively and qualitatively understandable by linking its most important topics and concepts to existing knowledge bases like Wikipedia or customer-specific ontologies, as well as to the underlying social networks. Its prime commercial use areas will include market research, business intelligence, election forecasting, emergency response and management, topic stickiness or popularity prediction, and trend identification.
The need for services enabling extraction of the most important topics and concepts from data and linking them to existing knowledge bases has increased due to the ever-accelerating volume and speed of social and online media, and due to the emergence of the so-called “fake news” phenomena. The fast speed of online communications requires prompt reactions if we want our actions to be effective. On the other hand, social and online media provides an easily accessed and rich source of text data for analysis and predictions.
The Innovation Activity Leader, Sarunas Girdzijauskas, Associate Professor at Swedish KTH Royal Institute of Technology said:
'The SenseMaking Innovation Activity will respond to demand by providing a web service that rapidly answers user requests regarding expected trending scenarios for given topics that are expressed as a set of input keywords. Our competitive advantage is the automatic and transparent topic extraction, allowing the linking of unstructured natural text to existing knowledge bases and to the underlying social graphs of web chatter. This allows the detection of causal social network communities and events, the integration of various and different knowledge bases, and comparison of data-driven and knowledge-based analysis in the same interface.'
A typical use case of the SenseMaking service is the prediction of the stickiness and the spread patterns of a given topic, such as a newly launched ad by a company or a newly started election campaign, expressed by the end user as a set of natural keywords in any language. SenseMaking performs rapid linking of the input data to up-to-date structured representations, and extracts the causal social network communities and events to return a prediction of the upcoming developments of the queried topic in the web chatter to the end-user, as well as recommendations on what could be done to further promote the topic or to contain its spread. Containment might be needed for topics related to emergency response situations, for instance, where the spread of rumours and fake information through web chatters might cause uncontrolled damage.
The SenseMaking service will provide information on the share of data covered by the respective topics, the sentiment towards the topic, as well as accurate predictions on how the topic might spread or stick in the near future. The user can easily analyse the data thanks to existing nomenclature defined in the knowledge bases, or explore the data in a purely data-driven fashion. The information is communicated to the user as text (the probability that a topic will stick, for example), and as graphs showing the expected spread.
Dr Magnus Sahlgren, the business champion of the SenseMaking activity, co-founder and Chief Scientist of Gavagai has said:
'The SenseMaking service can detect spikes of discussions on certain topics from online chatter and social media. It breaks them down into subtopics and detects sets of similar discussion spikes by using temporal topic similarity graph analysis. Further, it will provide an analysis of the underlying social networks which produced the events, such as news websites, politicians, blogs, web-trolls, and so on, depending on the issue. SenseMaking will be the tool to help decision makers make more sense from web chatter data, and predict the effects of social media on their campaigns, products, and so on, before it is too late to take action.'
Furthermore, with the help of machine learning techniques using features from the underlying social network and temporal topic-similarity graph made by the SenseMaking service itself, it can predict topic stickiness, and thus help, for example, to determine whether to take immediate action.
Innovation Activity Leader Dr Sarunas Girdzijauskas explained:
'At the first stage, our Sensemaking service will focus on Sweden and English-speaking countries. By the end of this year, we aim to have a fully functional SenseMaking service running at our SME partner and business champion Gavagai, and will consider transferring the intellectual property rights to a commercial stakeholder in 2018.'
The SenseMaking Service is one of the 13 innovation activities of the Digital Infrastructure Action Line of EIT Digital for 2017. In addition to the lead partner, KTH Royal Institute of Technology (Sweden), the other EIT Digital partners participating in the Innovation Activity are RISE SICS (Sweden) and the University of Trento (Italy). Gavagai, a Swedish language technology company and Sweden based communications agency United Minds supplement EIT Digital Partners by acting as Business Champions.
EIT Digital seeks to generate significant innovations from top European research. As such, we focus our investments on a limited number of innovation areas known as Innovation Action Lines, that we have selected with respect to their European relevance and leadership potential. Each Innovation Action Line comprises a portfolio of activities including:
- open Innovation Activities carried out by the EIT Digital Partners, and
- fast-growing technology startups that are ready to scale commercially.
EIT Digital Innovation Activities deliver new products or services, create startups and spinoffs to commercialise outputs from projects, and encourage the transfer of technologies for market entry.