# CDBC

**Repository Path**: liu-xiuzhen/CDBC

## Basic Information

- **Project Name**: CDBC
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: GPL-3.0
- **Default Branch**: master
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-12-07
- **Last Updated**: 2024-12-07

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# CDBC
Configuration/Diagnosis Benchmark in Choco

Knowledge base and datasets as configuration or diagnosis problems:

## 1- Knowledge Bases:
We have implemented Bike2KB, PCKB and CameraKB configuration knowledge bases in CSP using choco library in Java. 
### 1.1 - Bike2KB
Bike2KB has been generated based on the configuration benchmarks of IT-University of Copenhagen (https://www.itu.dk/research/cla/externals/clib/).
### 1.2 - PCKB
PCKB has been generated based on the configuration benchmarks of IT-University of Copenhagen (https://www.itu.dk/research/cla/externals/clib/).
### 1.3 - CameraKB
CameraKB has been generated by our project team based on real digital camera online markets.

## 2- Datasets:

### 2.1 - CameraKB_DiagnosisDataset:
In order to evaluate the success rate of the prediction quality of a diagnosis algorithm, we applied a user study with 264 students.  Users select values for the 10 different camera parameters (resolution, display size, etc.). Besides, users select 3 most important (weighted) parameters as w1=the first most important parameter, w2=the second most important parameter, w3=the third most important parameter. For these inconsistent user constraints, no product can be found in the product catalog. Then users need to select an available camera from the product catalog in the user study.  We use selected product ID to compare with the result of a diagnosis algorithm to evaluate its prediction quality.

### 2.2 - CameraKB_ConfigurationDataset:
We have converted the user requirements in CameraKB_DiagnosisDataset into consistent requirement sets.  Using this dataset, we can evaluate the prediction accuracy of a configuration system.