CSE498, Collaborative Design, Spring 2014
Computer Science and Engineering
Michigan State University

Urban Science is an analytics consultant to the automotive industry. Using data-backed analysis of key performance indicators (KPIs) Urban Science increases dealer sales and profitability.

A recent addition to Urban Science analytics is the Logic Tree, which describes important KPIs, target values for dealers, potential reasons for poor performance and potential suggestions to address problems. As Urban Science consultants learn from consulting experiences, the Logic Tree must grow and evolve.

Our Dealer Improvement Recommender System provides tools to visualize and edit the Logic Tree through a robust web app. Authenticated users can create, edit and delete KPIs, causes, suggestions and relations between them. Furthermore, these relations can be weighted to generate analytically driven suggestions based on a specific dealer’s data.

Urban Science’s existing Dealership Consultant Assistant iPad app is able to query the Logic Tree to provide actionable suggestions to consultants in the field. These suggestions are used to create dealer action plans. Results from tracking the implementation of these suggestions are used to update the Logic Tree, improve the system, and yield better long term suggestions for Urban Science’s dealers.

Our Dealer Improvement Recommender System is written in C# using ASP.NET MVC with a Microsoft SQL backend database accessed via the Entity Framework. Visualizations are implemented using D3.js.