ISYS1101 ISYS1102 Database Applications

23:59 Sunday 28 August 2022

System requirements
The system is developed in several phases. The first phase, which you are responsible for, is limited to
federal general elections for House of Representatives. The following voting processes are not in the
scope of this phase:
1. Federal general elections for senate
2. Federal by-elections
3. State and territory elections
4. City council and shire council elections
5. Referendums
6. Any other election services provided by AEC
In this assignment, you are required to analyse the database requirements, design the database
backend for the voting system, identify various database optimisations, and implement the system.
Elections
In Australia, federal elections are held every three years. In these elections, the voters get an
opportunity to vote for both lower house (the house of representatives) and upper house (the senate).
There are currently 151 seats in the lower house, and 76 seats in the senate.
As the scope of this assignment is limited to federal general elections for House of Representatives no
further details are provided on the upper house (the senate) and how the elections are conducted for
senate seats.
Members of the House of Representatives are elected by the voters registered in each electorate using
full preferential voting. Each electorate elects one member.
Electorates
For the House of Representatives, each state and territory is divided into electoral divisions (or
commonly known as electorates or seats). Population determines the number of electorates. To ensure
continued equal representation, the boundaries of these electorates have to be redrawn (redistributed)
periodically. As of last re-distribution based on 2017 population data, there are 151 electorates in
Australia.
State
Population (in
2017) Electorates
NSW 7,797,791 47
VIC 6,244,227 38
QLD 4,883,739 30
WA 2,567,788 16
SA 1,716,966 10
TAS 519,050 5
ACT 419,256 3
RMIT Classification: Trusted
RMIT Classification: Trusted
NT 247,512 2
Total 24,396,329 151
Election Process
When the sitting government is nearing its term (3 years) or under circumstances it is dissolved, the
Australian Electoral Commission (AEC) calls for nominations of candidates. Registered political parties
will then nominate their candidates for one or more electorates. Most political parties nominate
candidates for many electorates. Independent candidates can nominate themselves for the election.
Once the nomination process is over, AEC will determine the election date and will print ballot papers
for each electorate. A sample ballot paper is shown below. A screenshot of a real ballot paper (Higgins
electorate in 2016 election) also shown.
On the election date, registered voters are required to attend a polling station and cast their vote on a
ballot paper similar to above. The actual voting process is much more flexible with pre-poll voting,
postal voting, absentee voting, and declaration votes. However, for the scope of this assignment, we
only consider regular voting process on election day.
The preferential voting system
Candidates for the house of representatives are elected using the preferential voting system. In this
system, the voters are required to cast their order of preference to ALL candidates contesting in their
electorate. As shown in the above ballot paper, the voter has given first preference to the fifth
candidate on the ballot paper. The voter has given their second preference to the first candidate on the
ballot paper.
At the end of the election day, after all the polling stations are closed, the counting begins. The counting
of preferential votes is a complex process.
Step 1: Count of first preferences (primary vote)
In this step, all of the number “1” votes are counted for each candidate. If a candidate gets more than
half the total first preference votes, that candidate will be elected.
Step 2: Distribution of preferences
If no candidate has more than half of the votes, the candidate with the fewest votes is excluded. This
candidate’s votes are transferred to the candidates according to the second preferences of the voters
on the ballot papers for the excluded candidate. If still no candidate has more than half the votes, the
second-last candidate who now has the fewest votes are excluded and the votes are transferred
according to the next preference on the ballot papers. This process is continued until one candidate has
more than half the total number of valid votes.
This process is illustrated with a real example (distribution of preferences in Aston electorate in 2019
election) on the page 32 of the following document.
https://www.aec.gov.au/about_aec/Publications/electoral_pocketbook/2019/2019-electoral-
pocketbook.pdf
A screen shot of the above-mentioned page:
At the end of this two-stage counting process, the winning candidate is declared as the new Member of
Parliament for the corresponding electorate.
Components of the proposed system.
The proposed computerised election system must be able to conduct the entire election process for
federal house of representatives elections. The major tasks in the election process are listed below:
1. Maintenance of electoral role
2. Maintenance of information required for conduct of an election (such as basic election
information, electorate information, political party information, candidate information, etc.
Refer to details below.
3. Election Day – This process should mimic the manual process where a voter visits a polling
information, once identification is established a ballot paper issued, marks their preferences and
lodges the ballot paper.
4. Counting of ballot papers (counting is a complex process, refer to details below).
This system will maintain the following information.
1. Computerised Electoral Role
The system will maintain a computerised electoral role, i.e. a database of registered voters for each
electorate. For each registered voter, following information is stored:
• Title
• First name*
• Middle names (if any)
• Last name*
• Gender
• Date of Birth*
• Residential Address* (Unit number, street number, street name, suburb, postcode, state)
• (no letter box addresses accepted)
• Postal Address (as above, or can be different)
• Contact Details (daytime phone number, mobile phone number, email address)
• Electorate (determined by the system based on residential address)
There are around 17,259,000 Australians are currently enrolled to vote
(https://www.aec.gov.au/Enrolling_to_vote/Enrolment_stats/index.htm). The increase of the size of the
electoral role is approximately proportional to the population growth in Australia. The current
population growth in Australia is approximately 1.2%.
(https://population.gov.au/sites/population.gov.au/files/2022-04/2022-23_budget_overview.pdf)
2. Details of Elections
The following details about elections are required to be stored in the database.
• Election Serial Number (a unique code generated and stored by the system)
• Date of the election
• Type of election (house of representative, senate, by-election, etc)
• Total number of electorates
• Total number of registered voters (the number of registered voters at the closing of the
electoral role registrations for the corresponding election)
3. Details of Electorates
The following details about electorates are required to be stored in the database.
• Electorate Name (refer to page 156 – 158 of
https://www.aec.gov.au/about_aec/Publications/electoral_pocketbook/2019/2019-electoral-
pocketbook.pdf) for full list
• Name of the electorate
• Total number of currently registered voters
• Historical record of registered voters (the historical data are captured at closing date of the
electoral role registrations for the past elections. Both the date and no. of voters are stored)
• Name and party of the current member of parliament
4. Details of Political Parties
The following details about political parties are required to be stored in the database.
• Party Code (refer to page 160 of
https://www.aec.gov.au/about_aec/Publications/electoral_pocketbook/2019/2019-electoral-
pocketbook.pdf)
• Name of the party
• Party Logo
• Postal address of the party headquarters
• Secretary of the party
• Contact Person (name and other contact details such as daytime phone number, mobile, and
email)
5. Details of Candidates
The following details about political parties are required to be stored in the database.
• Name
• Political Party Code (or IND, if they are independent)
• Contact Details name and other contact details such as daytime phone number, mobile, and
email)
• Election Code
• Electorate Contesting
6. Computerised Ballot papers cast
The computerised ballot paper captures and stores voters’ preferences (similar to what’s written on a
paper-based ballot paper.
RMIT Classification: Trusted
RMIT Classification: Trusted
[Very important] To ensure integrity and confidentiality of the voting process, once a voter is issued a
computerised ballot paper, there should not have any identification records to positively identify who
cast that vote. As such, only the following data are stored with each computerised ballot paper.
• Election Code
• Electorate
• Preferences cast (i.e which candidate got the first preference, who got the second preference,
etc)
However, there must be a mechanism in place to record the issuance of a ballot paper to a voter. The
issuance record must capture the following information:
• Election code
• Electorate
• Polling Station Name
• Identifying details of the voter – these data should be sufficient to uniquely refer to a voter in
the electoral role
• Timestamp
7. Election results
At the end of counting process, for each electorate, the following result data are stored.
• Election Code
• Electorate
• Primary vote for each candidate (i.e. first preferences)
• Preferential vote count for each candidate, at the each iteration of elimination process (refer to
page 32 of https://www.aec.gov.au/about_aec/Publications/electoral_pocketbook/2019/2019-
electoral-pocketbook.pdf)
Assignment Tasks
Milestone 1:
You are required to build the data model for this application using an Entity-Relationship diagram. This
diagram should be developed on Oracle SQL Developer. This diagram should be of professional quality
with sufficient details that any other database personnel should be able to comprehend.
Then, convert your data model into the physical database design and finally generate the DDL script to
build the back-end database schema for the application. Ensure that primary keys and foreign keys are
correctly identified.
In order to complete this milestone, you are required to demonstrate your workings and the final DDL
script to your tutor during Week 5 lab sessions. It is very important to get it marked off by the tutor
before you proceed to Milestone 2. If your design does not meet the business requirements, your tutor
will allow you to make any amendments (only once) and present again for evaluation.
Milestone 2:
1) Identify the tables, in your final schema, that are expected to be extremely large and are expected to
grow over time.
For each of these tables:
• Describe the expected record size (in bytes or kilobytes), the estimated initial table size, and
estimated table size after 10 years of use.
• Describe a suitable storage strategy for such tables. Your answer must include the modified SQL
DDL statements to define your chosen strategy.
2) Identify 3 common queries that would need to be run frequently against the database (at least 2 must
include a join, and one must include a transaction with multiple steps).
For each of the queries:
• Produce the SQL to correctly produce the expected result.
• Identify what indexes would help. Identify the type of index and columns that are used to
build these indexes (justify your design).
• Show the SQL commands for building these indexes in Oracle.
• Show the query execution plans both before the index is added and after adding the index.
• Explain how the index was utilised (or not) and why. What join algorithms were used? What
changes would you need to make for the index to be properly utilised, or for a different join
algorithm to be used instead? (Provide concrete details of the changes).
3) Describe a suitable partition strategy for extremely large tables you identified in step 1. Include
details of the partitioning type and which columns/key should be used. You must include the SQL DDL
statements used to implement your partition strategy.
You must justify your design decisions. Include details about which of the above queries it will improve
the performance of and how it helps with concrete examples. (You must explain in clear terms – such as
partition pruning, partition joins, and parallel SQL, applicable to each of these queries.)
(Note: While you will be able to test that there are no syntax errors, you may not be able to actually
implement them on RMIT’s Oracle servers due to permission restrictions).
5) Before a voter is allowed to vote, to ensure the integrity of the election system, the system should
check if he/she had voted earlier on this election. Write a trigger – previouslyVoted() , to check if
the voter had voted before.
This trigger reads the election code, electorate, voter identification as inputs and returns a Boolean
value (true, if voted before and false, if not voted before).
RMIT Classification: Trusted
RMIT Classification: Trusted
6) Write a stored procedure – primaryVoteCount(), to complete the step 1 of the counting
process. This stored procedure requires election code and electorate name as inputs. It will read
Computerised Ballot Papers and does required processing, and update Election Results table with
primary votes (first preferences) received by each candidate in chosen electorate in the chosen election.
7) Write a stored procedure – distributePreferences(), to complete the step 2 of the
counting process. This stored procedure requires election code and electorate name as inputs. It will
read Computerised Ballot Papers and does required processing, and update Election Results table with
preference votes received by each candidate at each preference distribution in chosen electorate in the
chosen election.
Milestone 3:
You are required to do a demo of your complete application hosted on the school's Oracle server. These
demos will be conducted during the week of 5th September (Week 7). You should have some voter data
and stored in the application and the tutor will test out the functionality by using a test bed of data.
Tutors will use a standard testing plan for all submissions and you will only receive your second
milestone marks if you can demonstrate the functionality of your application in this milestone.