Using Data Mining Technique to Improve Billing System Performance  In Semiconductor Industry

Abstract

The new billing approaches are manly to apply the  integrated concept of data warehouse with relevant billing  data; in addition, use the methods of mining association rule  to sort out the Billing Quantities Pattern and then figure out  the billing quantities. Moreover, employ the Decision Tree  algorithm of data mining to find out the unit billing price. As  a result, the new billing approach is made of the methods of  data warehouse and date mining. This study is mainly focused  on improving the operation of current billing system to  establish the new functionality of the Billing quantities and  Billing price. As for the benefit of these two new functions, it is  not only able to lead into clients’ billing systems, but it is also  capable of upgrading the efficiency in rapid setup; especially  for the enterprises that already possessed billing system  internally but not yet implemented. In addition, it can also  reduce the difference in revenue, shorten the process of  issuing invoice, speed up the export operation, increase the  export efficiency and provide the revenue data for integrating  into the Executive Data System (EIS), Decision Support  System (DSS) and Business Intelligent System (BIS) to allow  enterprises making the right decisions promptly

Existing System 

In terms of the IC Testing Manufacturing, it is mainly  an industry that provides the Wafer and IC testing services  for clients in order to make them be able to differentiate  their products from different grades and increase the yield  by improving their manufacturing process to make products. So far, there are many IC Testing manufacturers still  apply the manual or semi-manual operation to handle their  billing data. In addition, it may not only establish a  complicate billing rule, but it also lacks of flexible variation  in the end. The main problem is consisted in the calculation  rules that are complicated with needing the flexible and  variable billing factors. However, operations of many  systems become even more complicated due to their lack of  flexibility, even if continuously adopt the variable billing  rules and factors, still, during the process of making  financial report and export, it will not only be short of the  quickness and correctness for operation, but it also needs to  invest even more manpower in checking the billing data.

Proposed System  

This study proposed applying the data warehouse and  data mining technologies to the billing system; in addition,  expected to establish the warehouse system for the billing,  utilize the data mining technology and analyze the billing  approach to obtain a high flexible billing function in order to promptly provide the revenue data for enterprises, speed  up the lead time of client’s billing patterns and improve the  efficiency in billing process; Thus, to achieve the expected contribution and response  to the demand for complicate and flexible billing system,  this study is seeking for completing the following purposes:  1. Establish the billing data warehouse:  To collect and categorize the relevant billing data and  historic data into a same platform by using the concept  and technology of the data warehouse in order to make  the billing analysis and calculation.  2. Establish the rules of the billing price and billing  quantity to setup the billing pattern:  3.Improve the billing efficiency in reducing the adjustable  time that needed to response to different billing  variation

CONCLUSION 

This study is mainly applied the data warehouse  technology to integrate the related billing information and  the billing qty pattern mining methods to figure out the  correct billing qty pattern and the billing quantities that  made the correctness rate of entire billing qty pattern to  upgrade up to 99% averagely. In addition, the billing  pattern is applied the data mining’s Decision Tree to find  out the most optimal unit price, and actually use its  Decision Tree generating rule to figure out the unit price,  and the correctness rates are all over 90% that improved by  23% than the current billing system. The changing time of  the billing quantities is changed from billing system’s one  week to 32 days and the improvement rate is 71%. The lead  time is reduced from one month to one week, and the  improvement rate is 76%. At last, used the real-time online  analysis technology of OLAP to build up a multidimensional,  multi-quantities combination intelligent  business system to provide for making decision analysis.

REFERENCES 

[1] Shu-Hsien Liao, Chih-Hao Wen, Pei-Yuan Hsian, Chien-Wen Li  and Che-Wei Hsu 2014. “Mining Customer Knowledge for a  Recommendation System in Convenience Stores”, International  Journal of Data Warehousing and Mining (pp. 55-86).

[2] W.H.Inmon, 2002,”Building the Data Warehouse”,John Wiley&  Sons,.

[3] K.Kimball,M.Ross and W.Thornthwaite, 1998,”The Data  Warehouse Lifecycle Toolkit,” NewYork: John Wiley & Sons,Inc.,.

[4] Hugh J. Watson, Barbara H. Wixom and Dale L. Goodhue ,2004.  “Data Warehousing: The 3M Experience, Organizational Data  Mining: Leveraging Enterprise Data Resources for Optimal  Performance (pp. 202-216).

[5] Akiko Campbell, Xiangbo Mao,, Jeddah, Saudi Arabia 􀄡􀄡  2017, ,”Multidimensional Business Benchmarking Analysis on Data  Warehouses”, International Journal of Data Warehousing and  Mining Volume 13 • Issue 1

[6] Chia-Chen Chen, An-Pin Chen, 2007 “Using data mining  technology to provide a recommendation service in the digital  library”, The Electronic Library, vol. 25, pp. 711.

[7] Chen, R. S., Chen, Yeh-cheng, and Chen, C. C. 2010,” Using data  mining technology to design an quality control system for  manufacturing industry”, Proceedings of European conference on  Computer science (pp. 272-276).

[8] Lior Rokach and Oded Maimon, 2008,,“Data Mining with Decision  Trees: Theory and Applications(Series in Machine Perception and  Artificial Intelligence)”, ISBN: 981-2771-719, World Scientific  Publishing Company.

[9] Kleissner,C.,”Data mining for the enterprise, ,1998,” IEEE  Proceedings of the 31st Annual Hawaii International Conference on  System Sciences, 7, 295-304.

[10] Damla Oguz, Baris Yildiz and Belgin Ergenc, 2013. “Matrix Based  Dynamic Itemset Mining Algorithm”, International Journal of Data  Warehousing and Mining (pp. 62-75).