Because We
Abstract
Major Goals: Major Goals: Main Project goals for TA1 1) Adaptive, scalable, generalizable local knowledge elicitation (a) Semi-automatically generate starter causal model from SME inputs and statistical text processing techniques, (b) Solicit latent local causal knowledge, unexpected factors, and non-empirical or implicit beliefs by adaptive narrative questions, (c) Generate the right questions and sample the right subpopulations with algorithms to optimize information gain per touchpoint.). 2) Specific, predictive computational models (a) Construct models that include hidden factors and explain conflicting causes; use exploratory clustering to identify and define local collective knowledge (b) Predict outcomes, collective uncertainty, and distribution across subpopulations by integrating Fuzzy Cognitive Mapping with collective models of knowledge. 3) Interactive decision-support and outcome exploration tools (a) Show why particular outcomes are predicted, through interactive decision trees and "what if' scenarios Main project goals for TA2. 1) Self-sustaining engagement for dynamic model updates (a) In collaboration with local NGOs, design and deploy a Community Information Exchange (CIE) Online Platform for both initial and ongoing data collection, based on our adaptive data collection platform used in the WASH domain, (b) Enable model updates by building local metaknowledge about likely disruptions, early indicators of change, and preferred alternatives into long-term system monitoring and model updating tools, (c) Transfer ownership of the predictive model to local groups for sustainment. 2) Immediate and continuing value tailored to characteristics of locals (a) Collaborate with local groups to align system modeling plans with local needs and priorities (i.e. understand area's current WASH situation). (b) Leave behind ways for individuals to access useful aggregate information (e.g. local maps of current water quality and uses of household water treatment).
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 30, 2022
- Accession Number
- AD1202506
Entities
People
- Alice Leung
Organizations
- RTX
- Tufts University
- University of Washington