Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Martin Chapman: Research Overview, 2019

36 views

Published on

King's College London

Published in: Science
  • Be the first to comment

  • Be the first to like this

Martin Chapman: Research Overview, 2019

  1. 1. Research Overview THuMP themes Dr. Martin Chapman King’s College London martin.chapman@kcl.ac.uk
  2. 2. Themes in THuMP 1. Software infrastructure. • Consult decision support system (DSS). 2. Human-machine interaction. • Consult chatbot. • Distributed Research Games (DRGs). 3. Resource allocation. • CAT (Trading Agent Competition, reversed) platform. 1
  3. 3. 1. Software Infrastructure. Consult: DSS to assist patients who have suffered a stroke to self-manage their treatment1 : Blood pressure (Nokia) Pulse (Garmin) Activity (Garmin) Heart Rate / ECG (VitalPatch) EHR (EMIS) Device Integration (Nokia) Device Integration (Garmin) Sensor-FHIR mapper EHR Integration (EMIS) EHR-FHIR mapper HAPI FHIR Message Passer Dialogue Manager Provenance Data Miner Argumentation Engine Drug Interactions Tablet Browser PC Browser Chat Server UI backend Sensor data Sensor data FHIR resources FHIR resources FHIR resources Processed patient data Patient data Fragment Processed patient data Processed patient data, goal Results Dialogue responses Interactions Data summaries Results Wearable EHR PIMS DNC Argumentation Focus: Connection of individual processing components, integration of data sources and user interaction. 1Martin Chapman et al. “Computational Argumentation-based Clinical Decision Support (Demonstration)”. Accepted, AAMAS, 2019. 2
  4. 4. 2. Human-machine interaction. 1. Consult: Implementing a chatbot to interact with patients. • Using argumentation to initially determine which types of interactions to support (e.g. information-seeking [4]). 2. Supporting the interaction of human participants with a game system (a DRG) in order to elicit network attack and defence strategies2 . 2Martin Chapman. “Cyber Hide-and-Seek”. PhD thesis. King’s College London, 2016. 3
  5. 5. 3. Resource Allocation. CAT platform: another DRG designed to evaluate user-designed markets by determining their attractiveness to autonomous traders. Contribution: introducing exchange of advice between traders3 . Problem: conflicting advice. Framing: who do I allocate my choice of market to? def reward ( trader , p r o f i t ) : rewards [ trader ] = p r o f i t def selectTrader ( ) : i f random ( ) > epsilon : return max( rewards ) else : return random ( traders ) Solution: ϵ-greedy. 3Martin Chapman et al. “Social networking and information diffusion in automated markets”. In: TADA-AMEC (AAMAS). 2012, pp. 1–14. 4
  6. 6. References Martin Chapman. “Cyber Hide-and-Seek”. PhD thesis. King’s College London, 2016. Martin Chapman et al. “Computational Argumentation-based Clinical Decision Support (Demonstration)”. Accepted, AAMAS, 2019. Martin Chapman et al. “Social networking and information diffusion in automated markets”. In: TADA-AMEC (AAMAS). 2012, pp. 1–14. Elizabeth I Sklar and M Q Azhar. “Argumentation-Based Dialogue Games for Shared Control in Human-Robot Systems”. In: Journal of Human-Robot Interaction 4.3 (2015), p. 120. 5

×