ICT Business Analysis & Data Visualization
ICT Assignment – Business Analysis Case Study (Stage I – 25%)
Report Submission Date: End of Week 8, Tuesday, 21st May 2019, 23:59pm
Presentation: Week 9 (during respective classes)
In this assignment, a frequently cited Boston housing dataset is given to you to apply data mining techniques to resolve a business problem. Assuming that you are taking a business analyst role in a real estate consulting firm, and you are asked to leverage your business analytical skills to make better use of the given housing dataset.
By finishing this assignment, you will achieve the following learning outcomes:
1. Review the current algorithms, methodologies and modelling for ICT business analytics;
2. Evaluate various ICT business analytic tools and techniques; and
3. Propose a business analytics report to solve practical problems identified in an ICT business project.
Main tasks to do in this assignment:
1. Understand the dataset
2. Relationships discovery among features
3. Potential analysis tasks that you could do to provide business suggestions
4. Justify the reasons why you target those potential tasks and the potential benefits that the expected results/finding/achievement could help to improve business
5. Present your work to the class
Task 1: Understand the dataset
Read the attached txt file “data_description.txt” that contains detailed description of the attributes in the given dataset. Use this file as dictionary to get the meaning of each attribute value. In your report, you need to present your initial analysis results of the dataset, including at least but not limited:
• the features (attributes) are included in this dataset, identifying the numeric features and nominal ones respectively;
• simple statistics summary, such as frequency, data distribution, maximum, minimum, average (mean), standard deviation, and etc.;
• any extra result(s) that you think can help to better understand the dataset; and
• any visualization diagram/chart/figure if necessary.
Task 2: Relationships discovery among features
We attempt to discover the relationships existed among all features (attributes). You have many choices to work out the hidden relationships. Here are some suggestions:
• at object level, to convert the raw dataset into a well-structured one in relational data model (hint: to apply normalization technique);
• at attribute level, to group attributes into smaller subsets with particular focuses based on your understanding;
• to work out correlations among all or selected attributes; or
• by any other way, to build up your own results.
Task 3: List any potential business analysis tasks with your justification
Based on your analysis and understanding so far to list a set of potential business analysis tasks that you can work on using the given dataset. You should consider some tasks whose results may bring business benefits, support strategy development, improve business process, and support decision-making. Your tasks can adopt any of the following data mining techniques:
• association rule mining;
• prediction; and
You need to provide strong reasons why you decide to do certain tasks, including the expected results/findings/achievement. Why do you think that your analysis could bring the benefits to improve the business.
Task 4: Presentation
You need to present your work before the other students in class. The presentation will be arranged in lab session in week 9. The content of your talk should cover everything you have done in the above tasks. Each presentation takes for about 10 – 15 minutes. In real world, you will have very short time to present your idea to customers to win business opportunity. You need to prepare PPT slides for our presentation.
Report to submit
You need to submit your report via Dropbox. You also need to submit your presentation PPT file. If you have any visualization file with large size, please submit the original image file in .jpg, .gif, or other format that can be opened by windows photo viewer. Your report should be saved in MS Word or PDF format.
A suggested report template could be:
Task 1: understanding of the database
Analysis result 1
Analysis result 2
Task 2: relationships among the features
Task 3: potential business analysis tasks
BA task 1
BA task 2