Logo Eurokleis        R e s e a r c h, I n n o v a t i o n  &  F i n a n c e

TrendMiner
The recent massive growth of online media and the rise of user-generated content (i.e. Blogs, Twitter, Facebook) requires a high ability to interpret written information in different languages​​, and in an efficient and convenient time. From the scientific point of view, new media pose different challenges due to the type of conversation that trigger: conversational, quick and short. Moreover, these conversations generate a deep-rooted stream of events within the context. Consequently, the pre-existing technological languages ​​are no longer adequate because they are not precise, scalable and portable.  

The aim of the project is to develop an innovative and open-source methodology to allow the synthesis in real time (summarisation) of multilingual information's flows generated by the new media.

The scope of the project will be achieved through an interdisciplinary approach that combines text processing methods (natural language processing), machine learning, economy and politics. 

A key novelty will be the related to the use of learning algorithms for the automatic discovery of new trends and correlations weakly supervised.

Accessibility and scalability will be ensured by a cloud computing infrastructure. The validity of the results will be evaluated through two high-level case studies: support to financial decisions (with the support of analysts, traders, regulators and economists) and monitoring of political opinions (with the support of politicians, economists and politician's journalists). 

The techniques used for the two case studies, then, can be used in many areas such as, for example, business intelligence and customer relationship management.

Eurokleis will adapt the learning algorithm to the finance and economy domain, and will measure the correlation between market movements and the existing opinions on the    financial instrument.

Through the support of international organizations and experts, and thanks to the specific expertise concerning the development and implementation of quantitative models based on artificial intelligence (Artificial Neural Networks) and agent-based models, Eurokleis will develop a web application which will display in real time the summary of the opinions regarding the selected financial instrument (sentiment). The web tool will also indicate the path of change in the short term, both in terms of pricing and volatility. This web application will be distributed free of charge, while the use of the underlying model can be sold to institutions interested in operating within financial markets.

Follow us on  
Follow us on
Follow us on YouTube Seguici su YouTube
Follow us on LinkedIn