Centre for Policy Modelling News

News from the Centre for Policy Modelling, including announcements of: new papers, special issues, books, workshops, projects, jobs and study opportunities.

09 June, 2016

Welcome to Shaheen Syed...

...who is a new Marie-Curie Early Stage Researcher, as part of the SAF21 ITN project.

Shaheen Syed

Thesis: Text Analytics for the 21st Century Fisheries


Since the 90s, it has been well-known that unstructured and semi-structured data constitute up to 90% of an organization’s data volume. The unprecedented growth of the Web and social media since then has only further increased the relative amount of unstructured and semi-structured data.

A great way to analyze this largely unstructured textual data is text analytics. Techniques such as sentiment analysis and named entity recognition are heavily being used in all sorts of research institutions and private companies. It enables the extraction of opinions on a given subject, create structured data from unstructured data, uncover other types of potential wealth and a lot more. It is a relatively new field of study and some amazing insights have already been found in e.g. Economics or Biology once they adopted text analytics.

My PhD research is aimed at implementing text analytics techniques into the fisheries domain as a whole. That is, we are investigating to what extent text analytics can be applied within the fisheries domain to gain more in-depth knowledge about fisheries and its e.g. stakeholders by utilizing quantitative computer science text mining techniques such as natural language processing and machine learning. A high degree of emphasis is placed on the investigation of different methods belonging to the same technique. This makes the various studies somewhat more explorative. However, some emphasis is placed on the predictive power of text analytics for the fisheries domain in the final stages of this PhD.

Slides from the workshop on the "Simulation of fisheries and coastal fisheries"

This workshop happened on the 6/7th June 2016, Manchester Metropolitan University.  The presentations were:
  • Anthony Charles (Saint Mary’s, Canada) “Fisheries as Systems” (slides)
  • Francois Bastardie (DTU, Denmark) “Towards holistic simulation of economic performance and impacts on fish stocks of alternative marine spatial plans with emphasis on fishermen micro-decision-making” (slides)
  • John Theodorou (TEI, Greece) “The massive fish kills effects to the Mediterranean coastal communities: The human responses in Greece to a Chatonella sp. toxic bloom (Maliakos Gulf, Aegean Sea) and to an anoxic upwelling episodes (Amvrakikos Gulf, Ionian Sea).” (slides)
  • Volker Grimm (UFZ, Germany) “Structural realism and theory development in agent-based models addressing practical problems” (slides)
  • Ernesto Carrella (Oxford, UK) “A generic fishery policy simulator” (slides)
  • Mike Bithel (Cambridge, UK) “Imposing fishing pressure data on a global individual-based ecological model” (slides)
  • Attila Lazar (Southampton, UK) “Assessing future environmental, livelihood and poverty changes in coastal Bangladesh: an integrated framework” (slides)
  • Richard Taylor (SEI, Oxford, UK) “Piloting ‘local’ fishery models with stakeholders on the South Kenya Coast. What did we learn?” (slides)
  • Steven Saul (ASU, US) “The interaction of fisher behaviour and fish population dynamics in the Gulf of Mexico: what can an agent-based model inform us about this relationship and its effect on stock assessment.
Information about the workshop website can be found at: http://ssfcm.wordpress.com

05 February, 2016

New Model: A model of making

This is a model of individuals constructing, taking apart, using tools, buying and selling objects – a model of making itself. The world of objects is an artificial world of 1D strings, but where there are complex affordances and constraints which make plans (the steps of how to make a particular object) are valuable.

The purpose of this model is to provide the simulation infrastructure needed in order to model the activity of making. That is individuals using resources they can find in their environment plus other things that other individuals might sell or give them, to design, construct and deconstruct items, some of which will be of direct use to themselves, some of which they might sell or give to others and some of which might be used as a tool to help in these activities. It explicitly represents plans and complex objects as separate entities in the model – embedding the “Atoms – Bits” distinction highlighted within the DiDIY project. This allows plans to be shared between agents which give the steps of how to make objects of use – either on a commercial or a free basis.

The framework is intended as a basis upon which many, more specific, models could be constructed, allowing the exploration of a variety of “what if” or counterfactual possibilities and thus give a concrete but dynamic and complex instantiation of the issues and situations discussed within the DiDIY project.

The model and its documentation are freely available at:

    http://www.openabm.org/model/4871

Some slides introducing the model to the DiDIY project available at:

    http://slideshare.net/BruceEdmonds/a-model-of-making

2 ABM PhD scholarships available

There are the following opportunities for phd Scholarships at the CPM in Manchester on the following topics:
Integrating Data Mining and Agent-based simulation – applied to Manchester “Troubled Family” data
http://www2.mmu.ac.uk/research/research-study/scholarships/2016/integrating-data-mining-and-agent-based-simulation--applied-to-manchester-troubled-family-data.php

Modelling the Dynamics of Town Centre Change
http://www2.mmu.ac.uk/research/research-study/scholarships/2016/modelling-the-dynamics-of-town-centre-change.php
This is a scholarship competition, so the positions will only give to the best students applying (regardless of the subject you apply for). For the process and conditions see http://www2.mmu.ac.uk/research/research-study/scholarships/

Deadline 9am GMT 21 March 2016.

17 October, 2015

New discussion paper: Culture trumps ethnicity! – Intra-generational cultural evolution and ethnocentrism in an artificial society

Culture trumps ethnicity! – Intra-generational cultural evolution and ethnocentrism in an artificial society

By David Hales and Bruce Edmonds

CPM discussion paper: CPM-15-256

Abstract. Ethnocentrism denotes behaviour and beliefs that are positive towards those who share the same ethnicity and negative towards others. Recent artificial society models have been interpreted as demonstrating how ethnocentrism might evolve under minimal assumptions. In these, evolution is modelled over generations of agents where new agents are born inheriting the ethnicity, behaviours and location of their parents. Behaviour does not change within generations but over many generations and agents only interact with their neighbours. We present a model that considers short-term cultural adaption, where agents may interact with any in a population and do not die or give birth but imitate and innovate their behaviours. While agents have a fixed ethnicity, they have the ability to form and join cultural groups and to change how they define their in-group based on both ethnic and cultural markers (or tags). We find that over a range of parameters cultural identity, rather than ethnocentrism, becomes the dominant way that agents identify their in-group producing high levels of positive interaction both within and between ethnicities. However, in some circumstances, cultural markers of group preference are supplemented by ethnic markers. In other words, whilst pure ethnocentrism (based only on ethnic identity only) is not sustained, groups that discriminate in terms of a combination of cultural and ethnic identities do occur. In these less common cases, high levels of ethnocentric behaviours evolve and persist – even though the ethnic markers are arbitrary and fixed – but they only emerge when combined with culture centric behaviour. Furthermore, cooperative ethnocentric groups do not emerge in the absence of cultural processes. The latter suggests the hypothesis that observed ethnocentrism in observed societies need not be the result of long-term historical processes based upon ethnic markers but could be more dependent upon short run cultural ones. We discuss these results as well as the dangers of over interpretation of models like this.

Paper at: http://cfpm.org/discussionpapers/152

Model at:  http://www.openabm.org/model/4744