The TargetView system
(There is also a Hebrew version of this writeup)
During the years 1998 and 2000, I was involved in a unique venture which aimed to establish in Israel a new bureau for risk assessment and credit scoring. This venture was known as Target Scoring.
The TargetView system was used in house as a tool for making many types of assessments based on disaggregative areas and the attributes the share. Obviously, such information is provided “in all probability”.
The LogIn screen was used to identify the employees.
The main screen to follow, compiled a summary based on many databases and statistical tables, some of which are produced by the Central Bureau of Statistics.
Before explaining the user interface, I will elaborate about the logic behind it. Groups of people are defined, for example, as disaggregative geographical areas, meaning that “area1” doesn’t have to be equal in any way to “area2”. They can have different size or population, but at the same time, they will represent a portion of the general population, having unique attributes such as wealth. Then, when a person is checked, the system makes assumptions based on the group he or she belongs to, and reflects these attributes on this person, without knowing him or her individually. This is useful for fraud detection (creating profiles of people and purchased), credit scoring (predicting financial strength and stability based on statistical data collected from other people in the same “group”, etc.).
The next step is to draft a Polygon on a given map, in order to display known and predicted (in all probability) data about this Polygon. The advantage of TargetView is with it’s ability to allow creating such polygons on the fly, and getting the information about them.
There are several methods provided for creating such polygons, among them entering a list of coordinates (“x” and “y” pairs), calculating an area based on a given radius, etc. The raw data used for the on-the-fly calculations looks like that:
Another utility was used for plotting points based on given data, on a map.
This system was very helpful for analyzing data and also for finding errors in analysis made by others…
No comments yet.