 |
 |
Solution For Telecommunication
Telecommunication enterprises have to deal with vast amounts of data due to their specific business transactions. The faster you can receive the data that your analysis is based on, the better are your chances to react upon fast moving trends. Trends are identified based on the actual situation of the market and are analyzed with handy data mining. Fast analyses are of vital importance especially in the sector of telecommunication, which is based on trends, that decide for or against bringing your products into the market.
Scope
The enterprise in this example produces several hundred million data records per day. These records contain information about the utilization of their customers, master data and outflow and inflow. In addition to that, every call and corresponding billing records are registered as well.. All this data is transferred to a datawarehouse and OLAP environment, where further analyses will take place. Data mining experts are working on ad hoc requests and prepare routine statistics about this data inventory. Because of the nature of the business and the trend orientated market often times ad hoc requests repeat and even become part of the routine statistics.
The main task is concentrated on optimizing these ad hoc requests and repeating analyses, since they are based on the not normalized and complex structure of the datawarehouse. Knowledge of how to optimize this process is the most critical factor for progressing.
Before using AlligatorSQL Enterprise Edition
With every ad hoc analysis, several files are created. Those include complex SQL scripts to retrieve data and Excel sheets to present the results. They are all saved in a local folder or into unstructured files on the file server. The datawarehouse “know-how” developed in multiple ad hoc analyses is spread over all computers and possibly still available in the brains of your data mining specialists, if they have good memory. However, obviously optimizing or even using the existing knowledge in this situation is almost impossible. Therefore, this results in redundant work and unnecessary time loss.
The Solution
In this case, it is also obvious what optimizing approach we should choose. The data miner has to be able to go back to existing knowledge. This way the data retrieval can be much more efficient, since the SQL scripts developed previously already contain the necessary data objects (tables and views). With aimed usage of AlligatorSQL Enterprise Edition data sources can be bundled and filtered, combined with the SQL scripts and than saved into separate projects. The resulting project can than be saved on the file server or in a knowledge management system. If similar analysis requests repeat, the data miner can go back to the saved project and he/she will automatically have access to the previously developed SQL scripts and files. Also all used database objects are bundled in one view.
After using AlligatorSQL Enterprise Edition
Conclusion
When using AlligatorSQL Enterprise Edition in datawarehouse environments, the knowledge of data mining experts can be preserved with the project management, which bundles the analysis data into projects. This approach saves costs of initial training of new employees and gives more efficient work on ad hoc and repeating data analysis. Recurrent queries can be saved into templates ( Adaptor ) and listed into a library. Those templates are presented to the whole team and can be extended on regular basis. The resulting knowledge database is the basis for all analyses.
