Travellersim

Slideshow Transcript

1. Slide 1: James Steiner Shawn Graham, Freelance Agent-Based Model Programmer Postdoctoral Research Fellow in Roman Archaeology http://www.turtlezero.com University of Manitoba http://home.cc.umanitoba.ca/~grahams TravellerSim: Studying site hierarchies and territories with agent-based modeling
2. Slide 2: The fuzziness of places leads to…
3. Slide 3: …dots on the map • Thiessen Polygons
4. Slide 4: Agent-based modeling • individuals – called agents - are simulated within a computer and allowed to interact with each other and with their environment agents environment
5. Slide 5: Characteristics of an ‘agent-based’ model o Each ‘agent’ is an individual, bounded, computing object. It is its own programme. o Each ‘agent’ is unique. o Every agent may have same suite of variables, but the distribution of individual values over the entire population of agents approximates a normal distribution. o Behavior depends on its unique combination of characteristics. Behavior can also be contingent on its location in the environment and its neighbors
6. Slide 6: The ‘environment’ o Agents move within a computing environment (in our case, a chessboard-like arrangement of computing agents called a ‘cellular automaton’) o The environment can be passive (mere ‘background’) o ‘Active’ environments can represent ecological systems (growing vegetation; forest fires; growth medium in a virtual petri dish) o The environment can also keep track of the impact of agents (each cell can register and record the presence/absence of agents, whether or not they have altered the environment (‘sheep’ agents eat ‘grass’ which must then ‘regrow’)
7. Slide 7: Rihll and Wilson’s 1991 model 1) Interaction between any two places is proportional to the size of the origin zone and the importance and distance from the origin zone of all other sites in the survey area, which compete as destination zones 2) The importance of a place is proportional to the interaction it attracts from other places 3) The size of a place is proportional to its importance
8. Slide 8: TravellerSim Interface • info
9. Slide 9: Travellersim Agents • Settlements – Passive, ‘genius locui’: accounting function; display importance by rescaling size • Travellers – Each has a ‘vision’ : knowledge of the world within average day’s journey – Tells each settlement where it has come from, ‘originally’
10. Slide 10: Where do I go from here? • Travellers examine their local area, select a destination on basis of ‘attractiveness’ • Attractiveness a function of – settlement’s importance (a function of the importance of the origin settlements of its visitors) – Number of visitors it has already attracted – Distance from current location of traveller • User controlled parameters: – Difficulty of communications – Benefit of concentrated resources
11. Slide 11: False false true false true
12. Slide 12: Social Networks Analysis • Fragmentation • Power • Flow-Betweeness • Degree • Factions
13. Slide 13: Sample Model Run
14. Slide 14: Model Outputs - Greece
15. Slide 15: Model Outputs - Italy
16. Slide 16: Focus on Factions - Italy
17. Slide 17: Extremely local geography – Tiber Valley Republican Farms and Villas farms villas starting map model run
18. Slide 18: Extremely local geography – Tiber Valley Republican Farms and Villas farms villas model run factions
19. Slide 19: Extremely local geography – Tiber Valley Republican Farms and Villas farms villas
20. Slide 20: Acknowledgements • Canadian Research Chair in Roman Archaeology, University of Manitoba • Mesa State College, Grand Junction Colorado • The British School at Rome