|
WizRule
Business Rules Detector
|
WizWhy
Analyzer and Predictor
|
WizRule reveals the rules governing the data and points out cases deviating from the discovered rules. The deviating cases having the highest degree of unlikelihood are listed as suspected errors or frauds. This method does not detect cases while fraud is done repeatedly because such a fraud creates a rule. WizRule meets this shortage by revealing the outstanding if-then rules. Such a rule is unexpected relative to others rules.WizRule reveals all the if-then rules, formula rules, and spelling regularities in the data under analysis. On the basis of the discovered rules, WizRule:
- points out deviations from the rules
- lists data entry errors – spelling errors
- reveals outstanding rules – unexpected rules
|
WizWhy determines how the values of one field(selected as dependent variable) are affected by the values of other fields (independent variable). The dependent variable is analyzed as Boolean. The user may select a non-Boolean analysis.WizWhy reveals all the if-then, if-then-not & if-and-only-if rules in the data under analysis. On the basis of the discovered rules, WizWhy:
- summarizes the data,
- points out interesting phenomena
- reveals necessary & sufficient conditions
- points out unexpected cases (cases to be audited)
- issues predictions for new cases – online or to a file
WizWhy applies the discovered rules (the model) to a new dataset (from the same population) and predicts the outcome for each record |
WizRule analyzes the data by revealing:
- Formula rules
- If-then rules
- Outstanding rules
- Spelling rules
|
WizWhy analyzes the data by revealing:
- all the if-then and if-then-not rules
- if-and-only-if rules (necessary & sufficient conditions)
|
WizRule issues the following reports:
- Overviw Report – display the value distribution in each of the fields providing quick data insights
- Rule Report – lists the discovered rules
- Spelling Report – lists cases that are likely to be spelling errors of names or other values (cases to be audited)
- Deviation Report – lists the cases that deviate from the set of the discovered rules (cases to be audited)
- Outstanding Report – lists unexpected rules to view interesting phenomena that may lead to a quick discovery of fraudulent cases
|
WizWhy issues the following reports:
- Rule Report – lists the discovered if-then & if-then-not rules
- Trend Report – presents graphically and textually the one condition trends in the data. WizWhy lists the relations between all the values in each field and the dependent variable.
- Unexpected Rules Report – lists the rules that are unexpected relative to basic rules and trends. These unexpected rules describe interesting phenomena in the data.
- If-and-only-if Rules Report – lists the necessary and sufficient conditions.
- Unexpected Cases Report – lists the cases (records) where the dependent variable’s value deviates from the expected value according to the discovered rules.
|
