INFS 5710 Information Technology Infrastructure for Business Analytics

Project Statement

(Due by 12 PM on Monday 2 August 2021 via Moodle)

• This project accounts for 30% of the total marks for this course. 

• The deliverable is a PowerPoint file with video narration and speaker notes.

Bike sharing has become increasingly popular across the globe. Today, such programs operate in roughly 1,000 cities, with more than half a million bicycles in use. The principle of bike sharing is simple: individuals use bicycles on an as-needed basis without the costs and responsibilities of bike ownership. It is short-term bicycle access, which provides its users with an environmentally friendly form of public transportation. This flexible scheme targets daily mobility and allows users to access public bicycles at unattended bike stations; bicycle reservations, pickup, and drop-off are all self-service. Commonly concentrated in urban settings, bike sharing programs also provide multiple bike station locations that enable users to pick up and return bicycles to different stations.

This project is about the bike-sharing scheme in the metropolitan area of Los Angeles (LA), a large city in the US with a population of several million. You are a business consultant working for the bike-sharing program.

Bike-sharing data

Your manager just referred you to download historical bike-sharing data from the following site: https://bikeshare.metro.net/about/data/, which contains data of more than 1.2 million bike trips from July 2016 Q3 – 2021 Q1.

You can also find in the data files the locations of bike stations in the GPS coordinate system. For example, the coordinate of a station is (x, y), where x is the longitude coordinate and y is the latitude coordinate. The following link helps you to understand more about the GPS coordinate system: 

https://www.ubergizmo.com/how-to/read-gps-coordinates/

Suppose a bike rental starts from (𝑥𝑥1 ,𝑦 1 ) and ends at (𝑥 2 ,𝑦 2 ), how should you estimate the distance traveled? In this project, it is recommended that you estimate it using the so-called taxicab distance, which is |𝑥 1 − 𝑥 2 | + |𝑦 1 − 𝑦 2 |. See the following figure for interpretation

For more information, please see https://study.com/academy/lesson/taxicab-geometry-history- formula.html.

Weather data

Weather plays an important role when people decide whether or not to use bike-sharing. You are required to explore the relationship between weather (e.g., temperature, wind speed and humidity) and the bike-sharing rentals in this project. Unfortunately, there is no easy way to download free historical weather data. The following provides a way to manually capture weather data month by month from Weather Underground (wunderground.com).

• First visit https://www.wunderground.com/ and try to search the weather condition in Los Angeles. (There are other locations that you may also try, e.g., Santa Monica and Valley Village, where there are many bike stations as well.) 

• You will be led to the site of a weather station near Los Angeles, which may be different from time to time.

 • Click the History tab on the page, and then choose to view Monthly weather data. Once you choose a month, click View. For example, the following link shows the weather data of June 2020 measured at the Burbank station (near Los Angeles): https://www.wunderground.com/history/monthly/us/ca/burbank/KBUR/date/2020-6

 • Scroll down the page, and you will see the table of Daily Observations. Use your mouse to copy the table and paste it to an Excel spreadsheet. 

• Copy only the data required, i.e., July 2016 Q3 – 2021 Q1, for this project.

Holiday data 

Another factor that influences the bike-sharing rentals is holidays. You can easily search the dates of federal holidays in the US for each year. The Task Your manager asked you to collect and analyze the data and “let the data speak.” You understand that the company wants to further grow the market and induce more users. Before they do it, they want to have some insights from the data. 

In this project, you are expected to manage and clean the data collected; some of them may contain missing data, different formatting, and incomplete information. The goal is to overcome such obstacles commonly encountered in reality to derive business insights from the datasets that can be used to promote the bike-sharing program in LA. The following are some aspects of analysis that you may consider:

• Station analysis: For example, what stations are most popular (for start or destination)? At what times?

 • Trip analysis: For example, what routes are most popular, one-way or round trip? What is the average distance of trips? Are most trips within a city or cross cities?

• Time analysis for demand: For example, what time or day has a higher demand? • Holiday analysis: How holidays affect the demand?

 • Weather analysis: How weather influences the demand? 

• Customer / Subscribe analysis: For example, what type of pass is most popular? Does it have any relation with the trips?

Regardless of the analysis topics that you choose, you must conduct a chronological analysis for each topic chosen. For example, how has the daily or weekly demand pattern evolved over the past few years? Most notably, COVID arrived in the US in early 2020. From your chronological analysis, you should be able to see how the pandemic impacted LA’s bike share program and people’s perception of biking. Also note the introduction of motor bikes in late 2018 (and COVID), which may have significantly changed customers’ demand for bike sharing. Basically, you should study pattern changes over years or quarters rather than present an overall, averaged result, per se. It is more preferred that you study one topic in depth than multiple ones superficially. Therefore, you are expected to study no more than three topics.

You are required to only use SAS Enterprise Guide (EG) for this project. To begin with the ETL (extract, transform and loading) process, you need to prepare your data in proper tables that will go into SAS. That is, you need to create tables in the SAS environment.

Whenever you want to conduct an analysis (e.g., trip analysis), you must write a query to select relevant attributes by properly joining multiple tables to obtain a resultant table for specific analysis. See Appendix for using some common data analysis functions of SAS EG. More features of SAS EG will be introduced in a tutorial session later.

Finally, please note that the management (or the LIC) does not know anything beyond this project statement. Therefore, you need to use your own judgement and make necessary and reasonable assumptions when doing this project. Make sure to present all assumptions made in the project.

Project Deliverable

Your group will submit a PowerPoint file with your video, audio narration recorded, and speaker notes. You should write your speaker notes in the Notes Pane for each slide. When you are recording your presentation video, you will speak following your own speaker notes in each slide. This will enable the LIC to both listen to your narration AND read your speaker notes when marking your project. 

DO NOT TURN IN A VIDEO FILE. PowerPoint includes a feature for recording slides. Here is a step-by- step reference: 

https://www.ou.edu/cas online/website/documents/Narrated%20Powerpoint%20(Office%20365).pdf Follow the steps for “Preparing to Record” and “Recording Narration.” You should ignore the last paragraph of this document on P. 4 and do not convert the PowerPoint file to a video file. 

Your presentation should be limited to 10 minutes with no more than 10 static slides that contain no animations or 'movement' of any description.

Slide structure 

The following is a suggested structure of your presentation:

 • 1st slide: Introducing group members 

• 1 - 2 slides: Briefly describe how you prepare the data for analysis, including how you clean data, manage missing information, and how you organize tables that go into SAS. 

• The rest slides: Topic I, Topic II, up to Topic III. For each topic, you should describe the research question, major findings (in terms of data visualization such as charts), business insights and recommendation. 

Marking guideline

Appendix: Using Enterprise Guide for Data Analysis and Visualisation

 Given a data file opened in SAS Enterprise Guide, you can see some analysis and visualisation functions available (from the tool bar below).

Histogram

If you are not familiar with the concept of histogram, please read the following site about histogram. To plot a histogram, choose Bar Chart Wizard. In Step 2 out of 4, choose Percentage for the Bar height. 

Correlation Analysis 

You may plot a 2D scatter chart first for the two variables that you want to study their correlation.

If a correlation is revealed from the scatter chart, you may also calculate the exact correlation between these two variables. Assume these two variables are “Amount” and “Visits”. The following figures show how their correlation can be calculated.

Drag Amount and Visits from the left pane to the right pane.