Picking the Winners: Forecasting Emergent Technology

This project brings together a unique combination of techniques from the research fields of Information fusion, bibliometrics/altmetrics, and topic modeling/text analytics in order to further advance the state-of-the-art within trend analytics / forecasting.

Indeed, the potential for following trends and perform forecasting has never been as viable as it is today due to the large number of different sources on the Internet. However, there is a tremendous challenge to utilize these huge and diverse streams of information for detection and prediction of patterns signaling shifts in trends which in the end could lead to new emergent technologies and innovations. The main expected result from this project is a framework, defined as a set of methods, that enables one to reliably detect and forecast emergent trends based on bibliometrics/altmetrics, topic modeling, and information fusion. One use case for the project will be forecasting of technology related to what is known as intelligent pharmaceuticals where the concern is to develop technology for effectively supporting the health care providers and patients to obtain optimal effects of medications. The aim is that such technology should delimit the significant global issue denoted by the World Health Organizations (WHO) (2013), as “non-adherence”. This relates to patients, for whatever reasons, not following long term treatments, leading to serious consequences both for their health and the society. 

Picture groupmembers PET

The picture shows the project participants (from left): Lovisa Österberg, Anders Hogner, Eric Rowcliffe, Staffan Truvé, Kerstin Forsberg, Alexander Karlsson och Nasrine Olson

Illustration project PET

The figure illustrates how so-called "Topics" evolves over time. Topics are extracted based on advanced data analysis (machine learning) performed on text data from various sources within an area of interest. As an example, one see from the figure that topic A splits into E and D and that E is a result of merging topics A, B and C. By following these topic flows, one can obtain an understanding of the current trends within the area.

Logo University of Borås

Logo Recorded Future

Logo Astra

KK-stiftelsen

Project leader

The project in short

Full name of the project: 
Picking the Winners: Forecasting Emergent Technology through Bibliometrics/Altmetrics, Topic Modeling and Information Fusion (PET)

Durance: 
2016-12-15 – 2018-12-14

Financed by: 
The Knowledge Foundation

Collaborating partners/Partnership: 
University of Borås, AstraZeneca, Recorded Future

Researchers related to the current project

Alexander Karlsson, Dr. (University of Skövde)

Mahboobeh Parsapoor, Dr. (University of Skövde)

Nasrine Olson, Dr. (University of Borås)

Staffan Truvé, Dr. (Recorded Future)

Daniel Langkilde, (Recorded Future)

Kerstin Forsberg, Lic. (AstraZeneca)

Lovisa Osterberg, Dr. (AstraZeneca)

Anders Hogner, Dr.  (AstraZeneca)

Eric Rowcliffe, Dr. (AstraZeneca)

Udo Bauer, Dr. (AstraZeneca)