IS428 AY2019-20T1 Assign Victor Lin Visualisations

VAST Challenge MC2: Citizen Science to the Rescue

Overview

Data

Interactive Visualisation

Observations & Findings

References

Interactive Dashboard

Dashboard Link: https://public.tableau.com/profile/victor.lin.27#!/vizhome/Dashboard_15709750608170/Intro

Homepage

This page contains the problem statement as highlighted by the case, as well as navigational buttons to each of the dashboards.The 3 dashboards that branch out from here are:Overview, Sensor Data, and Coverage All 3 dashboards have a navigational button at the bottom left corner to redirect the user back to this homepage.

Overview

Here the user will be able to see an overview of the sensor data, such as the missing gaps in some of the sensors' data, the spikes and elevated readings across the entire period. Although there is no interactive portion of the dashboard, it is meant to help the user see and understand the bigger picture.

Sensor Data

Through this dashboard, the user can filter out and see specific sensor data, such as the readings as well as the location. The multiple values checklist allows for more than one sensor to be filtered out for analysis. The readings chart plots the readings from the sensor across the 5 days, while the map shows the location of the sensor at the timestamp specified. The interactive portion of this dashboard comes from allowing the user to pick the sensor that requires analysis, as well as choosing a timestamp to see where in the city that particular sensor is at that time. Using this dashboard, the user is able to analyse the values on the readings chart and determine where the sensor was when it picked up higher/abnormal readings.

Coverage

This dashboard allows the user to select multiple sensors across specific time frames to do analysis. The multiple values checklist allows for more than one sensor to be filtered out for analysis, while the timestamp filter can be adjusted to reflect the time period of interest. This dashboard also allows for values to be filtered out, helping to sieve out readings which are unusually high. This is especially useful in determining regions in the city which may be contaminated. The dashboard can also be used to track which sensors were in the region within a time period, allowing the user to identify potentially contaminated cars.

Interactive Features

Technique	Rationale and Implementation
Navigation	Allow user to move between the dashboard pages with ease. Implementation: Create a dashboard sheet in tableau Under Objects, first select "Floating", then drag "Button" object to bottom left-hand corner of the dashboard In the popup, change button type to "Text Button", title to "Home", change the navigation to the "Intro" dashboard page and change the background colour and border colour under the "Format" field.
Filtering	Allow user to sieve out sensors of interest for more specific analysis while ignoring the rest of the sensors. Implementation Drag "SENSOR ID" from Dimensions to the Filter box Select the checkboxes for the sensors of interest, for those that are to be used in the analysis Go to the "SENSOR ID" pill under filters and click on "Show Filter"
Slicing	Allow user to set the range of values of interest for variables such as time and reading values. For example, the user can now set a time frame of interest to better understand what happened after the earthquake. Implementation: Drag "Timestamp" from Dimensions to the Filter box Select "Range of Dates" Drag the slider to match the start and end dates respectively. Alternatively, the dates and times can be manually keyed in Go to the "Timestamp" pill under filters and click on "Show Filter"
Paging	Allow user to see specific data at a specific time, as well as past geographical locations of mobile sensors prior to the selected timestamp. Implementation: Drag "Timestamp" from Dimensions to the Pages box Click on the "Timestamp" pill in the Pages box and select "Exact Date" Drag the slider to pick a specific time of interest or use the dropdown list