BUS 237 Lecture Notes - Lecture 8: Online Transaction Processing, Cluster Analysis, Data Mining

• School: Simon Fraser University
• Department: Business Administration
• Course: BUS 237
• Professor: Peter Tingling

Document Summary

Information overload: suffering in making decisions because of overabundance of information. Dirty data: problematic data in database (ex. 999-9999-9999 for phone number, 213 for customer age, etc. ) Missing values: a non-profit organization can process a donation without knowing the donor"s age but a data-mining application can"t. Inconsistent data: when an area code changes, the phone number is likely to change as well as the postal code. Wrong granularity: too fine: clickstream data measures every activity done by the mouse, many actions are irrelevant, not fine enough: data to be too broad and can not narrow down to data needed. Too much data: too many attributes, too many data points. Online transaction processing (oltp): collecting data electronically and processing the transactions online. Batch updating (updating at the end of day) Data resource challenge: not fully utilizing data as an asset. Online analytic processing (olap)/ decision support system (dsss): dynamic systems that focus on making the oltp-collected data useful for decision making.