Module 1
For Module 1, we started with Week 2, where we were introduced to the design process of data warehouses, the four step dimensional modeling process, and as well as the Balanced Scorecard. Data warehouses differ with databases in that it performs OLAP to analyze data over a period time, by performing complex joins, to try to make a business decision. Databases on the other hand, performs OLTP and is best used for transactional data for operational purposes. Data from databases are loaded into data warehouses by using ETL to extract data from the source database, transformed, and loaded into the data warehouse. The four steps of the dimensional modeling process includes selecting the business process to model, declaring the grain of the business case, choosing the dimension tables, and lastly identifying the attributes in a fact table. We also learned about the data cube as a representation of many attributes across its axis and the operations that can be performed including slice, dice, pivot, and drill down to manipulate the data to get a result. Lastly, we explored the benefits of keeping a Balanced Scorecard to measure performance and identify KPIs to achieve company goals by looking at four processes - customer, financial, internal process, and learning growth. A really successful example of a company using the Balanced Scorecard is when Southwest Airlines was able to achieve their goal by mapping objectives with each of the 4 processes and measuring how well they are doing for each objective.
During Week 3, we went in more depth in regards to advanced design for star schema and learned the roles of surrogate keys, junk dimension table, role playing dimension, and the three types of slowly changing dimension and we were able to actually implement of few of these for Assignment I. We also got a better understanding of the attributes that a fact table should contain and the different types of facts. We were also introduced to snowflake schema, which is a normalized dimension table. Between star schema and snowflake schema, a business can choose one over the other depending on their use case and as well as considerations for storage space. The second part of Week 3 discussed data quality analysis to perform data profiling in order to check data quality and allow better data integrity and governance.
For Week 4, we identified the importance of dashboards to quickly allow business users that previously have to rely on IT department to produce their reports, to create or understand information at a glance. By using visualization dashboards such as Tableau, it is much easier for users to find trends or look at major KPIs to assess their current business performance. One should try to avoid the pitfalls when creating a dashboard including having too many widgets, clustering of screen, exceeding the page boundary, or not having the context matching the data.
In conclusion, I have found Module 1's material to be extremely insightful and I know I am able to use many of the techniques I've learned and apply it to real business scenarios. Although I have found Assignment I to be very challenging as I have never encountered multiple time dimensions, I have learned alot from the assignment and can now identify the grain, facts, and dimension tables much quicker. Also, by coming up with questions on my own for the clinic use case, I got a better understanding of how data warehouses are tied directly to business goals. In regards to the Balanced Scorecard, I have learned that the key to having a successful Balanced Scorecard is constantly staying on top of it and use it like a living document. I also appreciate learning about it in this class because when the topic of Balanced Scorecard was brought up in my Business Foundations of IT course for the Continous Improvement part of ITIL, I knew exactly why the Balanced Scorecard would come in handy. Lastly, I do want to mention that I really appreciate the inclusion of Hans Rosling's video in the lecture as I found it so interesting and powerful at the same time. Rosling really explained to me that as human beings we tend to make certain presumptions that can totally be proven to be inaccurate through data and dashboards! I also enjoyed following his data points to see the different countries establishing a better health and lifestyle over time.
References
I. Kaplan, Robert S., and David Norton. Using the Balanced Scorecard as a Strategic Management System. Retrieved from https://hbr.org/2007/07/using-the-balanced-scorecard-as-a-strategic-management-system
II. Drkusic, Emil. Star Schema vs Snowflake Schema. Retrieved from
https://www.vertabelo.com/blog/data-warehouse-modeling-star-schema-vs-snowflake-schema/?fbclid=IwAR3SQHirvljHiZ_g6sFwflaaerw_TQGbYgOkYLo-WRJlhikPlNvrjZXgSRM
Thanks for reading!
During Week 3, we went in more depth in regards to advanced design for star schema and learned the roles of surrogate keys, junk dimension table, role playing dimension, and the three types of slowly changing dimension and we were able to actually implement of few of these for Assignment I. We also got a better understanding of the attributes that a fact table should contain and the different types of facts. We were also introduced to snowflake schema, which is a normalized dimension table. Between star schema and snowflake schema, a business can choose one over the other depending on their use case and as well as considerations for storage space. The second part of Week 3 discussed data quality analysis to perform data profiling in order to check data quality and allow better data integrity and governance.
For Week 4, we identified the importance of dashboards to quickly allow business users that previously have to rely on IT department to produce their reports, to create or understand information at a glance. By using visualization dashboards such as Tableau, it is much easier for users to find trends or look at major KPIs to assess their current business performance. One should try to avoid the pitfalls when creating a dashboard including having too many widgets, clustering of screen, exceeding the page boundary, or not having the context matching the data.
In conclusion, I have found Module 1's material to be extremely insightful and I know I am able to use many of the techniques I've learned and apply it to real business scenarios. Although I have found Assignment I to be very challenging as I have never encountered multiple time dimensions, I have learned alot from the assignment and can now identify the grain, facts, and dimension tables much quicker. Also, by coming up with questions on my own for the clinic use case, I got a better understanding of how data warehouses are tied directly to business goals. In regards to the Balanced Scorecard, I have learned that the key to having a successful Balanced Scorecard is constantly staying on top of it and use it like a living document. I also appreciate learning about it in this class because when the topic of Balanced Scorecard was brought up in my Business Foundations of IT course for the Continous Improvement part of ITIL, I knew exactly why the Balanced Scorecard would come in handy. Lastly, I do want to mention that I really appreciate the inclusion of Hans Rosling's video in the lecture as I found it so interesting and powerful at the same time. Rosling really explained to me that as human beings we tend to make certain presumptions that can totally be proven to be inaccurate through data and dashboards! I also enjoyed following his data points to see the different countries establishing a better health and lifestyle over time.
References
I. Kaplan, Robert S., and David Norton. Using the Balanced Scorecard as a Strategic Management System. Retrieved from https://hbr.org/2007/07/using-the-balanced-scorecard-as-a-strategic-management-system
II. Drkusic, Emil. Star Schema vs Snowflake Schema. Retrieved from
https://www.vertabelo.com/blog/data-warehouse-modeling-star-schema-vs-snowflake-schema/?fbclid=IwAR3SQHirvljHiZ_g6sFwflaaerw_TQGbYgOkYLo-WRJlhikPlNvrjZXgSRM
Thanks for reading!
Comments
Post a Comment