Business Analytics Case Study

Java Python Business Analytics Case Study

Case Description

The final business analytics case study has two components: a) data transformation with SSIS and  b)  predictive  analysis with RapidMiner. Both tasks are based on individual flight experiences from an airline company, aiming to improve customer satisfaction. Review the instructions below carefully before start the assignment.

This assignment is worth 40% of your assessment for the subject. This assignment is due at 16:00 PM on 25 October 2024.

PART I: Data transformation for customer satisfactory data with SSIS

In this part, students are required to set a data transformation process with for the relevant source data table and create target tables. A brief description of all source tables is:

•   Individual flight experiences arestored in “flight_ 1.csv ” and “flight_2.csv” .

•   Flight types are stored in “type.csv”, where the Type Code is derived from the first two letters of the Flight ID in “flight_ 1.csv ” and “flight_2.csv” .

•   Additional information about customers is available in “customer.csv” .

•   Addresses for all customers are stored in “address.csv” .

On this basis, perform. transformation and create:

1.   Fact table that contains all available measures related to individual flight experiences and this table must be connected to dimensional tables.

2.   Customer dimension table that contains information of customers.

3.   Time dimension table that contains information of time.

Students need to complete the template to describe the transformation process, which contains the following components:

1.   Transformation Process

1.1 Screenshot of the data flow page: Each of your screenshot must have “green ticks” and “number of rows displayed” .

1.2 Screenshot of the data viewer for the fact table

1.3 Screenshot of the data viewer for the customer dimension table

1.4 Screenshot of the data viewer for the time dimension table

2.   Challenges and Suggestions: Describe the data granularity and challenges related to the transformation process and make suggestions about how the company improve or make the data useful for business applications (max. 300 words).

Part II: Predictive analysis of customer satisfactory via RapidMiner

In this part, students must utilize RapidMiner to perform. predictive analyse on individual flight experiences from an airline company. The provided dataset contains various features that describe customer demographics, flight  characteristics, and service ratings. Students are required to develop a satisfaction prediction model (e.g., decision tree and logistic regression), and present it to the executives of the airline company who can make decisions whether to implement it or not. Based on the developed model, students need to complete the business report template where the goal is to improve/expand business operations. The report must contain the following components:

1.   Executive  Summary:  It is a clear, well-structured, evidence-based summary of your findings, which come from prediction performance from created models (max. 150 words).

2.   Predictive Analysis: Under the guidance of t Business Analytics Case Study utorials, Students are asked to develop two models (students can choose the models learned from this course or choose any other model outside the tutorial exercise) to predict whether the customer is satisficed for all customers in the provided “flight_test.csv” after learning patterns from the labelled “flight_train.csv” . Remember that you can switch between the different models by connecting the respective operator and temporarily disabling the operator that you do not wish to run at this moment.

2.1 Screenshots of Analytical Process for the first model: two screenshots are required, one for overall RapidMiner process and one for the process within Cross-Validation.

2.2 Screenshots  of  Analytical  Process  for  the  second  model:  two  screenshots  are required, one for overall RapidMiner process and one for the process within Cross- Validation.

3    Factors that influence  satisfaction most: regarding each model, identify and explain which variable(s) is/are the most decisive predictor(s) for customer satisfaction (max. 300 words in total). In this section, you need to insert screenshots of the related evidence to support your analysis for both models.

4    Classification Results: based on the classification results (confusion matrixes for both models), justify which model is better for the analytical task. You may also refer to related academic articles, blogs, and other publications to justify the modelselections by following the academic style. of writing and referencing (max. 200 words). In this section, you need to insert screenshots of the related evidence to support your analysis.

5    Suggestions  for improving company operations: propose specific actions on how to satisfy customers based on identified factors that influence satisfaction most for the selected model. You may also refer to related academic articles, blogs, and other publications to justify the model selections by following the academic style. of writing and referencing (max. 200 words).

6.   Ethical Problems: Discuss the ethical implications that you can identify from the data analytical process. You may also refer to  related academic articles, blogs, and  other publications to justify the model selections by following the academic style. of writing and referencing (max. 300 words).

7.   References: Please include all the references (minimum 5 references) that you have citied in your report.

Submission Guidelines:

This assignment is due at 16:00 PM on 25 October 2024. For late submissions, please refer to guidelines outlined on ECP:

https://course-profiles.uq.edu.au/course-profiles/BISM7233-60203-7460

You are required to submit an electronic version of your work through Blackboard (see the Assessment Link on Blackboard and look for the link labelled Assignment Submission. You will have two opportunities to submit your documents. So, in case you make a mistake on your first submission or wish to revise your assignment, you can re-submit it once. Also, by submitting the Assignment via Blackboard you attest that the Assignment is your own work.

Please submit the report on Blackboard Assignment Submission Box and Turnitin. You also need to submit the Zip file (compressed file) containing the SSIS project files, CSV output files, and RapidMiner Files on Blackboard Assignment Submission Box