Scientists seeking ways to solve complex natural resource management and conservation problems in poor nations often are reluctant to share their data fearing their work will be hijacked
A research team at the School of Informatics and Center for Genomics and Bioinformatics at Indiana University are constructing a model that seeks to build trust and encourage conservation scientists to make their findings public in a secure database environment.
Numerous biodiversity information systems have been developed and launched in recent years; however, data from those systems is not being shared or used to make conservation decisions in developing nations, the IU team claims.
Their findings are vital in helping policymakers in these countries balance the needs of their people with environmental concerns.
"Failure to understand the differences in the incentives for data sharing between scientists in developed and developing countries - where the world's richest biodiversity patches exist - is the very core symptom of the problem," said Sukamol Srikwan, a CGB conservation scientist.
"While professional publications provide a way to assign academic credit, and the notion that patents protect ownership of inventions, there is no similar system to protect raw data such as what a field biologist might observe and collect in a jungle".
Srikwan is the lead author of Trust Establishment in Data Sharing: An Incentive Model for Biodiversity.
She's joined in this ongoing study by School of Informatics researchers Markus Jakobsson, a computer security expert; and database specialists Andrew Albrecht and Mehmet Dalkilic.
The IU researchers are devising a system that allows field scientists to share their work with others similar to how much and videos are shared on websites, but with one important difference: they propose a system that establishes contributors' work as their own, using so-called cryptographic time-stamping techniques.
This approach makes it possible to prove in what order various documents were recorded, which in turn can be used to establish who 'owns' what data, thus allowing scientists to feel at ease when uploading their data.
"Our design has its foundation in computer security principles, but goes beyond the traditional approaches for access control and privacy," Jakobsson said.
"Many current data-sharing systems are designed by biologists with little or no guidance from computer experts - or designed by computer scientists with no input from biologists.
"Our work seeks to avoid these pitfalls".
The researchers now face the challenge of how to transform their proposed design into a working system and how and where best to deploy it, said Srikwan.
"Software engineering traditionally has ignored issues concerning culture, region, practices," Dalkilic said.
"This work embraces properties as just as significant in the design of a system".
The team will present their proposed model and other preliminary findings at the International Workshop on Trusted Collaboration in Atlanta, Ga.
The gathering includes other university researchers, scientists, industry professionals, software engineers and graduate students, and acquaints them with new theories and technologies related to security and privacy challenges in collaborative environments.
