Published January 1, 2012
| Version v1
Conference paper
Open
Data Mining Solutions for Local Municipalities
- 1. OLGU Comp Syst, Izmir, Turkey
- 2. Dokuz Eylul Univ, Dept Comp Engn, Izmir, Turkey
Description
This study proposes data mining solutions for local municipalities to make their decision support mechanism easier. The purpose of this study is to get intelligent solutions related to local government services from past data and to estimate the future activities. It covers socio-cultural analyses, income/expense analyses, infrastructure analyses, fraud detection analyses, simplification, verification and similarity analyses. Proposed system is based on service oriented architecture. The purposes of this project are; to give information about current state, to facilitate decision making for future activities, to increase income and decrease expense, to supply easy and correct data input to the system and to supply easier document tracking system. Seventeen scenarios were created initially. These scenarios are; Staff Analyzing, Classifying Citizens According to Real Estate Tax, Distribution of Citizens delaying Real Estate Tax, Income Operations Analyzing, Fuel Oil Analyzing, Electricity Consumption Analyzing, Cash Desk Analyzing, Distribution of Corporate Foundation, Moveable Material Analyzing, Logs Analyzing, Water Notice Analyzing, User Accounts Analyzing, Accountancy Analyzing, Employee Analyzing, Estimation of Wages, Citizen Analyzing and Corporate Foundation Analyzing. Service Oriented Architecture (SOA) is used as software architecture. Five services - Association Rule Mining Web Service (ARMWS), Outlier Detection Analysis Web Service (ODAWS), Classification Web Service (CWS), Clustering Web Service (ClustWS) and Data Preparation Web Service (DPWS) - were created. 7 scenarios used ARMWS, 3 scenarios used ODAWS, 2 scenarios used CWS and ClustWS is used by 5 scenarios.
Files
bib-194a1c3b-98b7-4345-abce-d558f64b1b98.txt
Files
(197 Bytes)
| Name | Size | Download all |
|---|---|---|
|
md5:4b1cedec2c208860684efe84f2da60a0
|
197 Bytes | Preview Download |