DAIS ITA: Year 1 Highlights and Annual Bootcamp
9 May 2018
The CSRI are working in partnership with the US Army Research Laboratory (ARL) and the UK Defence Science and Technology Laboratory (Dstl) as members of the Distributed Analytics and Information Science International Technology Alliance (DAIS ITA).
We have created an innovative approach to model the evolution of human groups on a generative basis, to anticipate their possible mutation. This approach enables exploration of tensions associated with affiliation to multiple groups and the influence on inclusion and exclusion of individuals. The insights could be useful in tense or unstable situations involving a range of human groups.
In collaboration with Yale University and Dstl, we defined a new mathematical model for the mutation of social groups, with a focus on group fracture, where individuals leave groups of which they are members. These models incorporate a selection of social and psychological theory which describes these phenomena as a way to understand their interplay and describe the trade-offs and challenges.
Colleagues at PSU and University of Massachusetts, Amherst developed a new approach to analyzing social network behavior. Using motifs, as originally proposed in biology, it is possible to analyze interactions in human networks that are too complex for traditional structural social network analysis. The concept is being explored further in the context of reciprocity with the team at Cardiff.
We worked with IBM to pioneer the use of a structured associative memory models called vector symbolic architectures, for representing and orchestrating complex decentralized workflows. This approach allowed use of artificial intelligence technologies to enable rapid self-organization and composition of complex services in a distributed manner.
Together with our partners at IBM UK and the US Army Research Laboratory, we created a new approach to exploit human-machine coalitions for situational understanding based on Subjective Bayesian Networks (Bayesian Networks with imprecise probabilities). The method learns effectively from sparse data and, coupled with conversational interfaces, can be used in real-world coalition decision making.
Led by Cardiff, researchers from BAE Systems, Cardiff and IBM created an experimental testbed for integrating an ensemble of machine learning and reasoning services - distributed across a coalition - to support decision-making informed by multimodal data sources and multiple modes of explanation.
DAIS ITA annual summer Boot Camp, 2018
The Distributed Analytics and Information Sciences International Technology Alliance (DAIS ITA) will hold its second annual “Boot Camp” meeting in June near Windsor, UK. Around ten members of the CSRI team will attend the three-day meeting to collaborate with peers from the other US and UK DAIS ITA partners. The Boot Camp will include planning for the programme’s Annual Fall Meeting to be held in upstate New York at IBM’s Armonk HQ in September.
For additional details on the research programme please visit the DAIS ITA Science Library, which contains all currently published publications, authors and organizations involved.