Indigenous Knowledge Aware Drought Monitoring, Forecasting and Prediction using Deep Learning Techniques Kidane W. Degefa kidane1982@gmail.com [or] kidane.woldemariyam@haramaya.edu.et Lecturer and Researcher at Haramaya Univeristy (Ethiopia) April, 2020
2 Drought is a natural environmental hazard causing adverse impacts on vegetation, animals, and people. 1. COMMUNITY ORIENTED SOLUTION (INDIGENOUS KNOWLEDGE) 2. TECHNOLOGICALLY ASSISTED SOLUTION (AI / DEEP LEARNING)
Statement of the Problem 3 Community Oriented Solution Certainty Structured Representation Technologically Assisted Solution Large data set requirement Model Interpretability & Visualization THE NATURAL PROGRESSION : STRUCTURED INDIGENOUS KNOWLEDGE BASED LEARNING
4 The general objective of this proposed research work is to design hybrid comprehensive framework for drought monitoring, forecasting and prediction using scientific and indigenous knowledge. KNOWLEDGE ORIENTED EXPLAINABLE- MODEL
Comprehensive Architecture 5 Knowledge Graph based Deep Learning Outputs Inputs Indigenous Knowledge aware Drought Different types of data Monitoring, Forecasting sources for drought and Prediction Models modelling Embedding Oriented Deep Learning Aggregation with Knowledge Graph Knowledge Graph Visualization based Explanation (Experts) (Non-Experts) Visualization/Explanation Drought Indigenous Knowledge Graph(Ontology)
What is new? 6 Learning from reasonable dataset Participatory Technological Solution Indigenous knowledge modelling and preservation AI model performance improvement and explainability Disambiguating and recognizing entities in context of drought
Conclusion 7 Deep Learning(connectionist AI) + Indigenous KGs(symbolic AI) = (Comprehensive, Explainable and Adaptable AI) [for Drought Monitoring, Forecasting and Prediction]
8 Today’s effective drought monitoring is Tomorrow’s life saver! Thank You for Your Attention!!!
Recommend
More recommend