9 areas of innovation
In peace informatics, we define nine areas of ‘data-driven innovation’, which are based on two dimensions: the stage of the ‘data-use-chain’ on the one hand and level of automation on the other. The data-use-chain includes collection, analysis and the dispatching of data streams. This is independent on the type of data used (e.g. whether we look at actively or passively collected data). The level of automation includes (groups of) individuals, algorithms and the crowd.
These distinctions lead to nine separate, interlinked areas of innovation. Each of these can be addressed in isolation, yet most would need to be seen as mutually dependent parts of changing data usage routines.
People-based data collection, such as household surveys. Innovation in this area includes digital technologies, for example with the introduction of handheld computers.
People-based data analysis, such as interpreting written documents and/or data sources. Innovative methodologies include text-mining or automated visualisation that helps to digest larger amounts of data more quickly.
People-based data dispatch happens when analysis results are being shared with those who are expected to make use of the provided information, for example when a researcher presents his/her findings to an expert policy-makers. Innovation in this area can come in the shape of social media platforms such as twitter, facebook, etc.
Algorithm-based data collection makes use of computational power to look for and merge available data sources (open or closed). Examples of innovation in this area include automated scraping methodologies of news data sources (e.g. GDELTproject.org)
Algorithm-based data analysis apply computational methods to give meaning to existing (often merged) datasets. Innovation examples include the growing methods aimed at automated sentiment analysis of social media content.
Algorithm-based dispatch refers to the technological possibilities that are entering the scene for automatically matching those who hold relevant information to those who are in need of it. The innovative space in this area revolves around systems aimed at for example early warning and emergency response.
Crowd-based data collection makes use of the fact that people are increasingly connected and able/willing to provide information to data platforms (be it actively or passively). Trends in innovation include local SMS-based trusted respondents’ networks that collect data on e.g. perceptions on public security.
Crowd-based data analysis leverages the capability of the crowd to look through large quantities of unstructured data streams (e.g. you tube, twitter, satellite imagery, citizen reporting) and to provide an interpretation; this can be done across large distances or with a clear link to local communities acting as the crowd.
Crowd-based data dispatch makes use of the crowd’s ability to decide where and when information should be distributed to. This could be done in the form of community feedback loops directed from the crowd towards affected communities (e.g. by sms messages).