Fast Algorithms for Mining Association Rules in Large Databases
VLDB '94 Proceedings of the 20th International Conference on Very Large Data Bases
A framework for designing the balanced supply chain scorecard
European Journal of Information Systems - Special section: PACIS 2004
A review of performance measurement: Towards performance management
Computers in Industry
Predictive collaborative performance system in B2B supply chain using neuro-fuzzy
ICOSSSE'10 Proceedings of the 9th WSEAS international conference on System science and simulation in engineering
| Hi-index | 0.00 |
Presently, modern B2B supply chain management (B2B-SCM) equipped with semi-automated data logging systems accumulate large volumes. However, each SC unit in B2B-SC still individually develops their performance. Besides, their linkages of performance attributes between SC units still lack vital information extraction to improve theirs. Therefore, this paper aims to propose an integrated framework between B2B supply chain (B2B-SC) performance evaluation systems and data mining techniques, developing relationship rules of collaborative performance attribute enhancement. The methodology is as follows. Firstly, B2B-SC performance evaluation questionnaires based on two levels able to characterize collaborative relation between two or more partners in their SC were gathered from the case study companies. The data set of relationships between enterprise and its direct customers of the case study companies in France was used for demonstration. Secondly, data cleaning and preparations for rule extraction were performed on the questionnaire database. The significance of attribute was calculated using attribute ranking algorithms by means of information gain based on ranker search. These results were used to choose the crucial attributes from each micro view. Thirdly, web graph analysis was performed on this data to confirm the strong attribute relationship. Next, association rule was deployed to extract performance attribute relationship rules grounded on support and confidence cross validation method. The quality of each recognized rule is tested and, from numerous rules, only those that are statistically very strong and contain vital information are selected. Last but not least, these rules are interpreted by domain experts and studied by domain engineers to build a collaborative performance attribute enhancement road map. Furthermore, the final rule set of extracted rules contains very interesting information relating to SCs and also point out the critical existing SC attribute improvement. Ultimately, companies in this SC are able to use this framework to design and adjust their units to conform with the exact customer needs.