Can computers be made to share knowledge like people do?

A new PhD project at Aarhus University will map how knowledge-sharing among large groups of people (crowdsourcing) works and then create a program enabling a computer to simulate human behaviour.

2013.06.10 | Line Ziegler Laursen

Oana Marilena Vuculescu

Knowledge-sharing among large groups of people – also known as crowdsourcing – has lately become something of a buzzword. One branch of crowdsourcing called Collective Problem Solving (CPS) enables large groups of people who do not usually work together to produce impressive solutions to complex problems.

In these big-data and knowledge-sharing times, the advantages of CPS are pretty obvious and have spurred PhD student Oana Marilena Vuculescu from the Department of Business Administration at Aarhus University to establish why and how this form of crowdsourcing works. She then aims to develop a computer model that can simulate human behaviour.

“I hope to design an agent-based computer model using so-called agents with specific coded attributes that enable them to simulate different types of people with different abilities. If the project is successful, the computer model will be able to identify the ideal constellation of people and knowledge for solving a given problem,” she explains.

New research approach
Numerous crowdsourcing studies have been conducted over the years, but Oana Marilena Vuculescu intends to adopt a new approach.

“Many studies focus on cases. However, because crowdsourcing is a complex system, it’s hard to determine how it works if you only look at the outcome. That's why I’ve chosen a different approach,” says Oana Marilena Vuculescu.

She started her PhD project in February 2013 and is still working on the preliminary phase, but expects to have produced a functional computer model capable of simulating CPS by the end of 2013.

Further information:
Oana Marilena Vuculescu, PhD student
Aarhus University, School of Business and Social Sciences
Department of Business Administration

Research news