Detecting Communities of Animal Trade Based on Google’s PageRank

Sep 8, 2026

Network analysis supports the study of animal trade by revealing patterns in movement networks and enabling community-based modeling. The work implements and tests a community detection algorithm aligned with an animal production zone definition, using Mato Grosso cattle-movement data for 2007. The dataset includes 15,844,764 animals moved. The algorithm incorporates directed and weighted links via a LinkRank adaptation of Google’s PageRank, producing 14 cohesive, geographically adjacent communities and identifying border counties that function as trade boundaries. Community analysis can inform prediction, survey planning, and veterinary preventive controls, especially when combined with GIS to summarize large livestock movement databases.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages1

Reading Time3 min

AccessFree to download

Community detection in animal trade networks

LinkRank-based method

Data analysis for Mato Grosso (2007)

Results: communities and trade borders

Applications with GIS for decision-making

Poster topic 08 Poster 29 Detecting communities of animal trade based on Google’s PageRank Grisi-Filho, J.H.H.1, Amaku, M.1, Ferreira, F.1, Dias, R.A.1, Telles, E.O.1, Ferreira Neto, J.S.1 and Negreiros, R.L.1,2, 1 University of São Paulo, Preventive Veterinary Medicine and Animal Health, Brazil, 2Instituto de Defesa Agropecuária do Estado do Mato Grosso, Brazil; [grisi@vps.fmvz.usp.br](grisi@vps.fmvz.usp.br) Network analysis has become a useful tool to study animal trade. Recent studies applied community detection analysis in animal movement networks, to improve predictive modelling and to help us understand the trade patterns of the animal production industry. We implemented and tested a community detection algorithm that reflects the definition of an animal production zone, providing useful information for researchers and decision-makers, specially to understand animal flow and identify trade regions within an area. We analyzed the State of Mato Grosso database of cattle movements for the year of 2007, provided by INDEA (Instituto de Defesa Agropecuária do Estado do Mato Grosso) . This database helds information about 15,844,764 animals moved. We implemented an algorithm that accounts for links direction and weight. It is based on the calculation of LinkRank, a modified concept of Google’s PageRank. In this sense, the community is defined as a group of nodes in which a random walker is more likely to stay. We think that this community definition is adequate to find regions of animal production, assuming that a community is a group of premises or counties that an animal is more likely to stay during its life. We found 14 cohesive communities, all of them containing counties that are geographically adjacent to each other. A small group of counties remained in the border of two or more communities, acting as trade borders. Community analysis is an useful tool to understand animal trade within a given area, and therefore to help predictive modelling or the planning of surveys and other veterinary preventive control measures. Also, network community analysis along with Geographic Information Systems can synthesize information in a very revealing way, avoiding one to be overwhelmed by the large amount of information maintained in livestock movement databases. ISVEE – Book of Abstracts 2012 383

FAQs

What is the document’s main goal in studying animal trade?

To detect communities in animal trade networks that reflect animal production zones and help explain trade flow patterns for decision-making and research.

How does the proposed method relate to Google’s PageRank?

It uses a modified LinkRank concept based on Google’s PageRank to define communities, accounting for directed and weighted network links.

What were the main findings from the Mato Grosso cattle movement data?

The analysis produced 14 cohesive communities, each with geographically adjacent counties, while small county groups on borders between communities acted as trade boundaries.

Sep 8, 2026

Upload by Theodore