Healthcare Fraud Prevention - Partner Database Appliance - Module Architecture

Key Information

Register
Submit
The challenge is finished.

Challenge Overview

Overview

The Healthcare Fraud Prevention Partnership (HFPP) is building a network in which public and private insurers will exchange data in order to detect fraudulent claims. Each partner in the network is an insurance company. The partners each have their own database of insurance claims.

TopCoder has implemented the data exchange network and is now working toward an end-to-end demonstration of its capabilities. In this contest, we are working on a module of prototype software that will represent a network partner. We have access to a set of claims like the ones collected by real partners, and we want to build a database appliance that stores and returns such claims.

 

Requirements

The database appliance will consist of a Python client supported by a relational database.

The relational database will store the sample claims provided by the Centers for Medicare and Medicaid (CMS):

There are five types of claim files:

  • beneficiary summaries
  • carrier claims
  • inpatient claims
  • outpatient claims
  • prescription drug events

The claim files are provided in comma-separated value (CSV) format.

The Python client must provide bidirectional access to the database:

  • given claims in CSV format, write them into the database
  • given a query for claims, read them from the database and return CSV files

The expectation is that we will start with an empty database and load all of the sample claims into it.

Subsequently, the database appliance will play a role in demonstrating the functionality of the data exchange network. Requests for claims will be passed through a network, and the database appliance will be called upon to provide the requested data.



Final Submission Guidelines

Software Environment

  • operating system: Linux (a recent version of Ubuntu or CentOS)
  • language of implementation: Python 3.3
  • relational database: any open-source relational database
  • we would like to rely on native Python libraries as much as possible
  • if you wish to use a third-party library, please ask the copilot in the forum or by using Contact Managers
  • all software must be freely available for unrestricted commercial use

 

Deliverables

  • standard deliverables for module architecture
  • we expect a direct-to-assembly approach to be completed with a single assembly contest

ELIGIBLE EVENTS:

2014 TopCoder(R) Open

REVIEW STYLE:

Final Review:

Community Review Board

Approval:

User Sign-Off

SHARE:

ID: 30035695