ABSTRACT

A good leaders or Government is the basic need to develop country. In India, who is largest democratic country in the world people are not fully involved in the selection process of Leaders. On an average there are 60-65% voting is done. For this purpose we have build the Data Warehouse that containing all the information related to election to increase awareness of voting. Using this we can find interesting patterns that are extracted and represented using Visual Data mining to arrange awareness program. The approach is divided into 5 phases: I) Data Preprocessing: II) Data Warehouse Creation: III) Task-Relevant Data Extraction; IV) Data Mining and V) Visualization. Generally, data collected from various cities of a state may be noisy, inconsistent and requires cleaning hence the database is preprocessed for missing values, normalization etc. in first phase. In Data Warehouse Creation phase, warehouse with vote as measure in fact and voter Gender, voter age, voter education, candidate, Religion, time, session, word as dimension. Word dimension has 4-level concept hierarchy of country, state, city, area/word id. Time dimension has concept hierarchy with year, Quarter, month, and date. Similarly, candidate has 2 level and Voter age, voter education, Religion, and session have level 1 concept hierarchy. In Last phase, results of data mining phase are represented using Visual Data mining techniques.