Cupid Minions

PROPOSAL

POSTER

APPLICATION

RESEARCH PAPER

Problem and Motivation

Low marriage rate and increasing divorce rate has been prominent issues in the developed world. One of the reasons for causing these issues is the wrong selection of marriage partner. Some couples only know that they don’t suit each other after marriage. This creates social issues like increaing divorce rate and domestic violence. Therefore, choosing the correct marriage partner is a prompt need for the young adults nowadays. In order to increase the successful matching rate for couples and make sure people can get the most satisfied partners, there is a need to understand these partner-seeking people’s characterstics, demographics, habits and lifestyle informations, etc. From the insight brought by the data, we could suggest the factors which contribute to successful and rapid matching and as a result help the partner-seekers.

Objective

The objective of the project is to:

• Understand the demographics of drivers
  1. Distribution of age of drivers.
• Understand the demographics of casualties
  1. Distribution of age of casualties.
  2. Distribution of severity of casualties.
  3. Distribution of type of casualties.
• Examine the underlying factors which contributes to accidents. The following are some factors, but not limited to:
  1. Temporal patterns: Accident records based on time.
  2. Weather conditions: Which type of weather conditions would cause more accidents?
  3. Road conditions: Which type of road conditions would cause more accidents?
  4. Location: Which city has the most accidents?
• Develop appropriate interactive visualisation to allow discovery of insights from multiple dimensions from the dataset.

Data

In this project, our team will be focusing on 2015 road safety data. The data is obtained from data.gov.uk (https://data.gov.uk/dataset/road-accidents-safety-data). It contains only personal injury accidents on public roads that are reported to the police and recorded using the UK STATS19 accident reporting form. It consists a total of 3 datasets that provide information about the accidents, the types of vehicles involved and the demographics of the casualties. Most of the data attributes are coded and re-coding would be done with the lookup tables provided by data.gov.uk.

The following data attributes are used in this project:

• Accident Dataset [Accidents_2015.csv]
  1. Accident_Index [Accident No.]
  2. Longitude
  3. Latitude
  4. Day_of_Week
  5. Time
  6. Local_Authority_(District) [City Name]
  7. Weather_Conditions
  8. Road_Surface_Conditions
• Vehicles Dataset [Vehicles_2015.csv]
  1. Accident_Index [Accident No.]
  2. Vehicle_Reference [Vehicle No.]
  3. Age_of_Driver
• Casualties Dataset [Casualties_2015.csv]
  1. Accident_Index [Accident No.]
  2. Vehicle_Reference [Vehicle No.]
  3. Casualty_Reference [Casualty No.]
  4. Casualty_Class [Driver/Rider, Passenger or Pedestrian]
  5. Age_of_Casualty
  6. Casualty_Severity

Research Visualisation

Visualisations	Comments
	Interactive Visualisation to Track Fatal Accidents (http://news.bbc.co.uk/2/hi/in_depth/uk/2009/crash/8414354.stm) The pie-chart on the left allows you to select the desired category of data to display. The radial bar chart will then allow you to look at the distinctive pattern of each age group over a time-period. The radial bar chart will be beneficial for high number of bins, where we will be able to look at all the bars or columns from one view without scrolling back and forth.
	Interactive Visualization to Rush Hour Danger (http://www.bbc.co.uk/news/uk-15975564) Visualization of the pattern of pedestrian casualties across the week Map which shows the deaths involving pedestrians and buses on London’s world-famous Oxford Street between 1999 and 2010 Although this visualization is interesting with the data points of casualties over the years, it lacks the interactivity that our team envision our storyboard to be. We will use this visualization as a reference when we implement an interactive map in our storyboard.

Tools

Following tools are expected to be utilized through this project:

• Tableau DeskTop 10
• JMP Pro
• Javascript
• D3.js
• JQuery
• Brackets
• Github
• Excel

Technical Challenges

Technical Challenges	Action Plan
Data Preprocessing - 195 Variables & 8378 Datasets	Collaborative team work in data cleaning, selection and transformation.
Lack of technical background in programming languages like, javascript and libraries like, D3.js and JQuery.	Initial hands-on experience during D3.js workshop Peer learning and sharing of skills developed during IS 480 Arranging frequent consultation with Instructor Prakash regarding technical difficulties encountered along the way
Unfamiliar with implementing interactive visual analytics application	Self-learning and practicing via online tutorials Continuous exploration of readily available alternative tools

Roles & Milestones