MAIN BOOKS 1 DATA WAREHOUSING IN THE REAL

MAIN BOOKS 1. DATA WAREHOUSING IN THE REAL WORLD : Sam Anshory & Dennis Murray, Pearson 2. DATA MINING CONCEPTS AND TECHNIQUES : Jiawei Han & Micheline Kamber, Morgan Kaufmann BUILDING THE DATA WAREHOUSE, W. H. Inmon, John Wiley & Sons DATA MINING TECHNIQUES : Arun Pujari, Universities-Press 3. 4.

ADDITIONAL READINGS INTRODUCTION TO DATA MINING WITH CASE STUDIES, G. K. Gupta (PHI) p DATA MINING Introductory & Advanced Topics, Margaret Dunham, Prentice Hall p

OLTP & OLAP p OLTP covers day-to-day operations of an organization such as purchasing, inventory, payroll etc. p OLAP caters to knowledge workers and aid in decision making

OLTP & OLAP p OLTP is customer oriented & used by clerks, clients & IT professional p OLAP is market oriented used for data analysis by knowledge workers including managers, executives and analysts

OLTP & OLAP p OLTP uses current data, usually very detailed p OLAP uses large amounts of historical data with facilities for summarization, aggregation. Information is maintained at different levels of granularity

OLTP & OLAP p OLTP uses E-R model p OLAP uses STAR or SNOWFLAKE model & a subject oriented database design

OLTP & OLAP p OLTP consist of small, atomic transactions. No of records usually in terms of tens p OLAP consist mainly of read-only operations. No of records accessed usually in terms of millions

OLTP & OLAP p In OLTP number of users in terms of thousands, DB size is 100 MB to GB, high performance, high availability Metric for measurement is transaction throughput p In OLAP number of users in terms of hundreds, 100 GB to TB, high flexibility & end-user autonomy Metric for measurement is query throughput

OLTP & OLAP p OLTP - unit of work is short simple transaction, detailed flat relational view, index/hashing on primary key for operation p OLAP – unit of work is complex query, summarized multi-dimensional view, lots of scan involved in operation