Vulnerability modeling in social care

We are currently developing a model building module that would greatly facilitate building the underlying reasoning model using multiple data sets and expert opinions.

Health and social programs represent a significant — and growing — proportion of the public sector budget across OECD economies. In fact, a recent report by OECD reported that public sector budgets devoted to social programs across OECD economies represented  22% of GDP.


Our vulnerability dashboard consists of a set of vulnerability metrics across several categories

Underlying technology

We use risk models (Bayesian networks) and dynamic models (Markov chain or Markov sets) which together provide the vulnerability metrics. The models can be created from a mixture of (i) individual historical data (ii) experts opinion and (iii) census (aggregated) information.

During 2013, we developed the underlying mathematical framework for multi-dimensional vulnerability assessment with limited information. We are currently applying the model on a real-world setting through the BJAST FOAK and possibly other use cases and we will develop methods to facilitate the construction of the underlying vulnerability model so as to make vulnerability dashboards easier and faster to deploy.


Lea Deleris

Example of results

Elder care dashboard

elder care dashboard