ACC7011
Individual Assignment – Business Analytics
Case study using software applications
40% of module marks
Case study
Crime is a complex social problem that affects all countries around the world. In recent years,
there has been a growing trend of crime in many countries, including Canada, Denmark,
Germany, Luxembourg, and Wales. You have been invited to a discussion forum to examine
the growing level of crime in these five countries, to identify some of the factors contributing
to this trend, and to discuss some of the challenges and opportunities for addressing this
issue.
Crime Rates in Canada, Denmark, Germany, Luxembourg, and Wales
During the discussion, the main organizer John Flynn presented fascinating statistics
regarding crime in all the five countries. According to Statistics Canada, the overall crime
rate in Canada increased by 5% in 2021, following a 3% drop in 2020. The most common
types of crimes in Canada are non-violent offenses, such as drugs and fraud. However, there
has also been an increase in violent crimes, such as kidnap and homicide, in recent years.
In Denmark, the crime rate has also been increasing in recent years. According to the Danish
police, the number of reported crimes increased by 4% in 2021. The most common types of
crimes in Denmark are property crimes, such as theft and burglary. However, there has also
been an increase in violent crimes, such as assault and kidnap, in recent years.
The crime rate in Germany has also been increasing in recent years. According to the German
Federal Criminal Police Office, the number of reported crimes increased by 1% in 2021. The
most common types of crimes in Germany are property crimes, such as theft and burglary.
However, there has also been an increase in violent crimes, such as murder and extortion, in
recent years.
The crime rate in Luxembourg has also been increasing in recent years. According to the
Luxembourg Police, the number of reported crimes increased by 5% in 2021. The most
common types of crimes in Luxembourg are property crimes, such as theft and burglary.
However, there has also been an increase in violent crimes, such as assault and murder, in
recent years.
The crime rate in Wales has also been increasing in recent years. According to the Welsh
Government, the number of reported crimes increased by 2% in 2021. The most common
types of crimes in Wales are property crimes, such as theft and burglary. However, there has
also been an increase in violent crimes, such as murder and drugs-related crimes, in recent
years.
Flynn then discussed a number of challenges and opportunities for addressing the growing
level of crime in Canada, Denmark, Germany, Luxembourg, and Wales. He first focussed on
the need for a comprehensive approach. This approach should include measures to address
the root causes of crime, such as poverty, inequality, and social exclusion. It should also
include measures to prevent crime, such as early intervention programs and community
policing.
Flynn also stressed on the need for cooperation: Crime is a transnational problem that
requires cooperation between countries to address. This cooperation is essential for
combating organized crime and other forms of transnational crime.
Finally, Flynn discussed the need for evidence-based policies: Crime prevention and control
policies should be based on evidence of what works. This means evaluating the effectiveness
of different programs and approaches and making changes based on the evidence.
Finally, he argued that technology can be used to prevent and control crime.
At this time, the authorities are interested in finding out the number of times specific words
are used in all the five countries to get a rough idea of how these crimes are discussed in
social media. For this purpose, they have collected all social media posts related to those
countries in specific files and are interested to know how the words with possible relation to
crimes are being used in these countries.
In relation to fraud, Flynn is also planning to build a model to help identify fraud before it
happens or as it happens, especially in credit card transactions. His intention is to build a
model to help identify fraudulent credit card transactions before they take place.
In this context, you are required to address the tasks below. Carefully follow the submission
guidelines provided at the end of this case. The datafiles are also available at the following
GitHub page: https://github.com/bibekbhatta/BusinessAnalytics within folder Assignment3
Requirement:
1) Present a screenshot of a table that shows the average number of times the words
'drug', ‘kidnap', 'murder', 'extortion' and 'fraud' have been used in each of the
five countries. Relevant files are provided to you on a folder called
“Crime_socialmedia_Files”. Also show the number of years of data available for
each of the countries in the same table. Submit your code in good form with
comments properly marked. 40 Marks
Marking Scheme:
Screenshot of summary statistics – 40 Marks (an excellent summary would include all of the
components required in the task in tabular form)
Failure to provide working codes will nullify the respective marks above.
2) Present a screenshot of a graph showing the number of times any one of the five
words (in 1 above) have been used over the years in all the five countries. Label your
graph appropriately. Submit your code in good form with comments properly
marked. 40 Marks
Marking Scheme:
Screenshot of graph – 40 Marks ( an excellent graph would include title [5 marks], labels in
axes [5 marks], graphs of all five countries clearly shown [20 marks ], and legend [5 marks]
in a well-presented [5 marks] format)
Failure to provide working codes will nullify the respective marks above.
3) Flynn provides you a file called “creditcard_fraud.zip” which contains data related to
credit card fraud. The last column ‘Status’ is a dummy variable of 1 if the transaction
is fraudulent, 0 otherwise. Use either RandomForest or DecisionTree to build a
machine learning model to predict the transactions that are fraudulent. For this
purpose, use all the columns of variables provided and use 70% of the data as a
training set and specifically use random state of 38. Show a screenshot of the
confusion matrix [10 marks] and provide a manual calculation of accuracy score [10
marks] of the model that you have used. 20 Marks
Marking Scheme:
Confusion matrix – 10 Marks (an excellent matrix would be easily readable with the required
information).
Calculation of accuracy – 10 Marks (calculation should be shown clearly)
Failure to provide working codes will nullify the respective marks above.
Submission guidelines
You should upload the following files in Canvas by the submission deadline:
• “ACC7011_answersheet.docx” (this answer sheet is provided in Canvas; do not change
the name of the file and do not edit the locked spaces)
All the codes that are required must be pasted properly within the given space in the provided
answer sheet ACC7011_answersheet.docx. Please do not submit files with any other names
other than specified above.
The codes provided by students should be replicable in Python 3; i.e. the codes should run in
a Python 3 environment without ANY issue except for the file path. Comments are to be
marked appropriately so that it does not interfere with the smooth running of the codes. If the
codes do not work, mark of 0 will be awarded for that part of the question. Do NOT change
the names of the files or the folders or the extensions of the files/folders.
Caution: The instructor might compare images, screenshots, text etc across submissions
using various techniques including artificial intelligence (AI), if needed.
Plagiarism is treated very seriously; click here for more information on plagiarism. For any
use of AI, guidelines provided in the module outline must be strictly followed.
If your ability to do the assignment is affected due to personal circumstances (e.g. health
issues, family issues, etc.), follow the guidelines provided on Exceptional Circumstances.
Any questions that may arise subsequently will be addressed in the FAQ page. You
should check this page regularly and prior to submission as well.
ACC7011
个人作业 -使用软件应用程序的商业分析案例研究
40%的模块标记
个案研究
犯罪是一个复杂的社会问题，影响到世界各国。近年来，包括加拿大、丹麦、德国、卢森堡和威
尔士在内的许多国家的犯罪率呈上升趋势。您应邀参加一个讨论论坛，研究这五个国家日益增长
的犯罪水平，确定导致这一趋势的一些因素，并讨论解决这一问题的一些挑战和机遇。
加拿大、丹麦、德国、卢森堡和威尔士的犯罪率
在讨论中，主要组织者约翰·弗林（JohnFlynn）介绍了有关所有五个国家犯罪的有趣统计数据。根
据加拿大统计局的数据，加拿大的整体犯罪率继 2020年下降 5%之后，2021年增加了 3%。加拿
大最常见的犯罪类型是非暴力犯罪，例如毒品和欺诈。然而，近年来，绑架和杀人等暴力犯罪也
有所增加。
在丹麦，犯罪率近年来也在上升。据丹麦警方称，2021年报告的犯罪数量增加了 4%。丹麦最常
见的犯罪类型是财产犯罪，例如盗窃和入室盗窃。然而，近年来，袭击和绑架等暴力犯罪也有所
增加。
近年来，德国的犯罪率也在上升。根据德国联邦刑事警察局的数据，2021年报告的犯罪数量增加
了 1%。德国最常见的犯罪类型是财产犯罪，例如盗窃和入室盗窃。然而，近年来，谋杀和敲诈勒
索等暴力犯罪也有所增加。
近年来，卢森堡的犯罪率也在上升。根据卢森堡警方的数据，2021年报告的犯罪数量增加了 5%
。卢森堡最常见的犯罪类型是财产犯罪，例如盗窃和入室盗窃。然而，近年来，袭击和谋杀等暴
力犯罪也有所增加。
近年来，威尔士的犯罪率也在上升。根据威尔士政府的数据，2021年报告的犯罪数量增加了 2%
。威尔士最常见的犯罪类型是财产犯罪，例如盗窃和入室盗窃。然而，近年来，谋杀和毒品犯罪
等暴力犯罪也有所增加。
弗林随后讨论了解决加拿大、丹麦、德国、卢森堡和威尔士日益增长的犯罪率的一些挑战和机遇
。他首先着重谈到了采取综合办法的必要性。这种方法应包括解决犯罪根源的措施，如贫困、不
平等和社会排斥。它还应包括预防犯罪的措施，例如早期干预计划和社区警务。
弗林还强调了合作的必要性：犯罪是一个跨国问题，需要各国合作解决。这种合作对于打击有组
织犯罪和其他形式的跨国犯罪至关重要。
最后，弗林讨论了循证政策的必要性：犯罪预防和控制政策应基于有效的证据。这意味着评估不
同计划和方法的有效性，并根据证据进行更改。
最后，他认为技术可以用来预防和控制犯罪。
目前，当局有兴趣了解所有五个国家使用特定词语的次数，以大致了解社交媒体上如何讨论这些
罪行。为此，他们在特定文件中收集了与这些国家有关的所有社交媒体帖子，并有兴趣了解这些
国家如何使用可能与犯罪有关的词语。
在欺诈方面，弗林还计划建立一个模型，以帮助在欺诈发生之前或发生时识别欺诈，尤其是在信
用卡交易中。他的目的是建立一个模型，以帮助在欺诈性信用卡交易发生之前识别它们。
在这种情况下，您需要完成以下任务。请仔细遵循 本案例末尾提供的提交指南。
要求：
a) 展示一个表格的屏幕截图，该表格显示了“毒品”、“绑架”、“谋杀”、“勒索”和“欺诈”等词在
五个国家/地区中每个国家/地区的平均使用次数。相关文件在名为
“Crime_socialmedia_Files”。此外，在同一表格中显示每个国家/地区的可用数据年数。以
良好的形式提交您的代码
正确标记的注释。 40分
评分方案：
摘要统计的屏幕截图 – 40分（一个优秀的摘要将以表格形式包括任务中所需的所有组件）
未能提供工作代码将使上述相应标记无效。
b) 展示一张图表的屏幕截图，显示这五个单词中任何一个词多年来在所有五个国家/地区的
使用次数。适当地标记图形。以良好的形式提交您的代码，并正确标记注释。
40分计分制：
图表截图 – 40分（一个优秀的图表将包括标题 [5分]、轴上的标签 [5分]、所有五个单词的可区分
形式的图表 [20分]和图例 [5分]以清晰呈现的 [5分]格式）
未能提供工作代码将使上述相应标记无效。
c) 弗林为您提供了一个名为“creditcard_fraud.zip”的文件，其中包含与信用卡欺诈相关的数据
。如果交易是欺诈性的，则最后一列“状态”是一个虚拟变量 1，否则为 0。使用
RandomForest或 DecisionTree构建机器学习模型来预测欺诈易。为此，使用提供的所有变
量列，并使用 70%的数据作为训练集，并特别使用 38的随机状态。显示混淆矩阵的屏幕
截图 [10分]，并提供准确度分数的手动计算 [10
标记]。 20分
评分方案：
混淆矩阵 – 10分（一个优秀的矩阵很容易阅读所需的信息）。
精度计算 – 10分（计算应清楚地显示）
未能提供工作代码将使上述相应标记无效。
投稿须知
您应该在提交截止日期之前在 Canvas中上传以下文件：
• “ACC7011_answersheet.docx”（此答题纸在 Canvas中提供;请勿更改文件名，也不要编辑锁定的
空格）
所有必需的代码必须正确粘贴在提供的答题纸 ACC7011_answersheet.docx的给定空间内。请不要
提交上述名称以外的任何其他名称的文件。
学生提供的代码应该在 Python 3中可复制;即代码应该在 Python 3环境中运行，除了文件路径之外
没有任何问题。注释应适当标记，以免干扰代码的顺利运行。如果代码不起作用，则该部分问题
将获得 0分。请勿更改文件或文件夹的名称或文件/文件夹的扩展名。
注意：如果需要，教师可能会使用包括人工智能（AI）在内的各种技术比较提交的图像、屏幕截
图、文本等。
剽窃会受到非常严肃的对待;点击这里了解更多关于剽窃的信息。对于人工智能的任何使用，必须
严格遵守模块大纲中提供的指南。
如果您完成任务的能力因个人情况（例如健康问题、家庭问题等）而受到影响，请遵循特殊情况
中提供的指南。
随后可能出现的任何问题都将在常见问题解答页面中解决。您应该定期查看此页面，并在提交之
前查看此页面。