4: accept

Summary Evaluation: Paper Description: Explain briefly the paper's contribution

The paper defines requirements for representing geospatial entities and their relations, focusing on geographic needs. It surveys the literature and compares six existing agent-based toolkits.

Summary Evaluation: Strong Points: What are the paper's strong points (2-3 sentences)?

Reduction of the necessary relations to placement and neighbourhood, divided into four kinds. Systematic comparison of the studied ABM platforms.

Summary Evaluation: Weak Points: What are the paper's weak points (2-3 sentences)?

The work seems to be still in an initial stage. Detailed description of TerraME-ABM is not given.

Detailed Comments: Please supply detailed comments to back up your scores. These comments will be forwarded to the authors of the paper.

Paper well written, work well planned. The paper defines requirements for representing geospatial entities and their relations, focusing on geographic needs. It surveys the literature and compares six existing agent-based toolkits. The work seems to be still in an initial stage. Detailed description of TerraME-ABM is not given.

3: weak accept

Summary Evaluation: Paper Description: Explain briefly the paper's contribution

The paper describes agents and cells (geographic regions) as geospatial entities and discusses the need to consider the four possible connections between those entities (agent to agent, agent to cell, cell to cell and cell to agent). The connections are used to organize a table that compares six simulation frameworks.

Summary Evaluation: Strong Points: What are the paper's strong points (2-3 sentences)?

The immersion of agents into a geo-referenced environment adds a very relevant dimension to the social simulation research field. The comparison of agent based simulation toolkits, regarding its capability to deal with geographic information, is an useful contribution of this paper.

Summary Evaluation: Weak Points: What are the paper's weak points (2-3 sentences)?

The paper claims to contribute towards generalizing the representation of geospatial entities and their relations but it is not clear what is the achieved more general representation; it seems that the generalization is found along the lines of the agent to agent and the cell to agent relations but the paper is very descriptive about those relations whereas a more rigorous (formal) explanation would help to better clarify the approach. The paper lacks an illustrative case study to help sustaining the descriptive argumentation.

Detailed Comments: Please supply detailed comments to back up your scores. These comments will be forwarded to the authors of the paper.

The paper refers a previous work (from different authors) where the cell to cell and agent to cell were described within a leader-follower interpretation. The authors argue that the agent to agent and cell to agent are not inferred from the previous two relations but the argumentation does not provide an alternative interpretation (for the previous leader-follower). Additionally the proposed placement relations (section 3.1) seems to consider that the cell to agent is inferred (symmetric) from the agent to cell relation. The paper would benefit from a clearer characterization of the proposed extension from previous work.

The comparison of agent based simulation toolkits (table 1) is difficult to understand; for example it is not clear which column represents the agent to cell relation; some information in the table cells seems not to be explained elsewhere, e.g., transition; also the idea of complex relations is left to be explained.

A simple modeling example, using the 4 proposed relations, would help to evaluate the consequences (e.g., regarding the generalization capability) of using those relations in the simulation modeling process.