Study of Slowly Changing Dimension to Dependent Data Mart to Manage Finance Data

Financial institutions face many challenges of managing and marketing campaigns which leads in its data warehouse. The management of marketing campaign leads in dependent data mart with real time updating and recording difficulties especially when many campaigns are running parallel ways. To securing the customers from being contacted too often for sales-based marketing contacts, the concept of novelty skeleton are introduced to clamp the customers who have been targeted in Sales based campaign for a specified time period. During the novelty Frame, the customer cannot be targeted by other Sales based campaign categorized under the same channel. The introduction of novelty skeleton has increased the difficulties of campaign management and data management. The difficulties of data management include timely update and robust storage system of campaign leads. In this paper, we explained represent the concept of slowly changing dimension on dependent data mart and also studied how it can be used in the data mart of financial institutions to update and maintain marketing campaign records of customers.


I. INTRODUCTION
The Financial institutions generate huge volume of data daily due multiple transactions and operations.These data are analyzed by the financial institutions such as bank to provide valuable customer insights such as the customer behavior, needs and desires.
Unexpected ways of campaign and management may result in customer exasperation.Robust and wellmanaged data Mart that stores and tracks customer contacts history which can assist in effective campaign management.However, with the set of activities and presence of large amount and volume of data, the institutions face lot of difficulty to drive the dependent data mart.
We can think two main categories of campaigns in the financial institutions which are based on sales and non-sales in nature.
The sales based campaigns target selected set of customers.A campaign leads to market for selling financial institution's products and their services.
Non-sales based campaigns serve and provide 191 customers with fulfillment and notification services.
These campaigns are break into two types of medium for campaign distribution-calling and non-calling types [5].The calling types include telemarketer call, relationship manager call and branch officer calls.Non-calling types include direct mailer, Email blast, electronic direct mailer and short messages [5].
Analysts select and target the most appropriate segments of customers for marketing campaigns from the dependent data mart.Below figure shows that the source data goes to the data warehouse for storage purposes then it create the data mart.Source data for the dependent data mart is extract from operational systems as well as systems that may be external to the business unit for which the data mart constructed.Source data may also be extracted for Enterprise Data Warehouse• if any exists [4].
Customers who are targeted for many campaigns may be over contacted during a period of time which may take result in unhappy from the institution and may spread bad experience.Proper handling process is needed to ensure that the result does not take place unhappy experience.
A powerful data management process will rule an efficient data operating process to manage present and historical data over time in the data mart for any industry.
Slowly Changing Dimension (Kimball, 2008)  Type 1: Type 1 SCD overwrites old records with new records.Type 1 SCD is easy to maintain.However, no historical observations are kept in the data warehouse [5].
Type 2: Type 2 SCD updates the record by inserting new observation while preserving the historical observations.Unlimited historical observations can be preserving with this type [5].
Type 3: Type 3 SCD updates the record by creating new dimensions to the table structure.It preserves history -only the previous record could be preserve [5].
The observation covers Type 1 SCD by updating the latest record.

Source data
Source data For non-calling campaign channels, the novelty skeleton is (T + X) days, where T is the lead time blast date and X is the lapse time period in days.For calling campaign channels, the novelty Frame is determined by successful contact of the leads.In the prospective where lead has not been successfully contacted, the novelty Frame will be (T + Y) days, where T is the lead time blast date and Y is the lapse time period in days.In the prospective where the lead has been successfully contacted, the novelty skeleton is (Z+Y) days, where Z is the lead contacted date and Y is the lapse time period in days [5].
The DCHT checks on the status of the leads.Below are a few prospective of the leads: (1) Fresh leads which was not targeted in any past campaigns (2) Already availing leads which was targeted in past campaign and are out of novelty skeleton (3) Already availing leads which was targeted in past campaign and are still within the novelty skeleton Existing channel of campaign differs from the newly targeted channel of campaign.
In prospective 1 and 2, where fresh leads which were not targeted in past campaigns and existing leads which were out of novelty skeleton will be successful uploaded onto DCHT.DCHT will only consist of actionable marketing leads.In prospective 3, where existing leads which were targeted in past campaigns and are still within the novelty skeleton will be categorized two sub-categories.Existing leads where present medium of campaign is different from the new targeted channel of campaign.These leads will be successful uploaded onto DCHT.Existing leads where present medium of campaign is similar to the new targeted medium of campaign.These leads will not be uploaded onto DCHT.This is because novelty skeleton is distinctive at channel level.Figure 3.1 is explaining the derivation of DCHT from multiple Analyst Tables.The contact histories of customers are remaining as unlimited historical observations can be secured with SCD type 2. After some time, the contact histories of customers are continuously built on a single database and may cause the database to spiral out of data space.

V. IMPROVEMENT
The SCD Type 2 is that the data is regularly built on a single type of database causing the database to expand fast way in overtime.My suggestion is that for diverting historical data onto an Archive

Figure 1 . 1
Figure 1.1 Dependent Data Mart is the name of a data management process that loads data into multiple dimensions of tables which contains data.To study and adopting Slowly Changing Dimension, the data has to change slowly on an irregular and random schedule.There are three kinds of Slowly Changing Dimension methods, Type 1, Type 2, and Type 3. The most commonly practiced SCD types are 1, 2 and 3. Below we present the descriptions of the different Types of SCD [5].

Figure
Figure 3.1

Figure 5 .
Figure 5.1 showing that the how improved data management are uploading leads onto the DCHT and Archive Table.And also Figure 5.1 shows the improved workflow of campaign lead and their selection process.

Figure 5 .
Figure 5.1 Improved Data Management

Id Customer Name Passport Number Effective Start Date Effective End Date
SCD Type 2 is relatively simple in implementation.The ability to retrieve current and historical data with Type 2 is also useful in reporting purpose.In view of that, SCD Type 2 is widely used in scenarios such as capturing customers' information and supplier's information etc.Customer Keynecessary for the financial institution like bank and insurance companies to implement marketing campaign novelty Rule for all sales-based campaigns purpose.The novelty skeletons will lock-on the customers who are targeted in Sales-based campaign for a time period known as the novelty skeletons.
Table and remaining current campaign in DCHT.Records that are 0 in the present variable/field will be archived into Archive Table.When Analysts load Analyst Table then the Archive Table will be updated onto DCHT.