INFS5700 Introduction to Business Analytics
Week 9: Big Data and Algorithm Ethics (T2 2022)
• Recap on Week 8 topic – Design Thinking
Copyright By PowCoder代写 加微信 powcoder
• Topic on ethical aspects of business analytics
Why Design Thinking?
• Paradigm shift from an ‘analytics world’ to a ‘business universe’ perspective
• Human-centred designs and outcome- oriented approach to big data projects and solutions
What is Design Thinking?
• Design thinking is a holistic approach that aims for innovation in a similar manner to the way designers would
• Design thinking is an iterative process in which we seek to understand the user, challenge assumptions, and redefine problems in an attempt to identify alternative strategies and solutions that might not be instantly apparent with our initial level of understanding
Design Thinking Process Flow Chart
Cumulative continuous improvement (KAISEN)
Persona Development (Anthropology)
End User persona aims to develop a rounded view of stakeholders (i.e.
people that will interact with your analytics use case) Diverge
Storyboard
A narrative storyline that aims at putting theory into practice scenarios (acting script) to identify potential problems and generate problem statements
Opportunity Matrix
Explore analytics opportunities with ideas generation that could translate into solutions.
Leverage Matrix (Prioritise)
Opportunity Canvas
Sort ideas and opportunities according to practical implementation/value creation and prioritise for prototyping and testing. Converge
Tools to help operationalise DT process for Analytics
• It is the foundation of a human-centred design process
• Theworkyoudotounderstandpeople,withinthe context of your design challenge
• Yourefforttounderstandthewaytheydothingsand why, their physical and emotional needs, how they think about the world, and what is meaningful to them
• The goal is to identify a grand challenge or problem faced by the focal company/organization
• It should be a guiding statement that focuses on insights and needs of a particular user, or composite character
• Togeneratedesignsolutionspertinenttothe challenge statement
• Itrepresentsaprocessof‘goingwide’with concepts and outcomes. It is concerned with ‘Flaring’ rather than ‘focus’
• Itprovidesalargerepositoryofdiverseideasthat are the source material for building prototypes to test with users
Persona Development
• To explore the problem in a grounded way, empathising with the customer and thinking about the problem from their perspective (e.g., understanding of users’ needs, experiences, behaviours, and goal)
• Creation of archetypes to represent a person’s demographics, behaviours, beliefs, motivations, intentions
• It should be developed from perspective of people in the system (i.e., both external users, e.g., customer, or internal users, e.g., staff)
Storyboarding
• It communicates a concept by visualising user interactions. They use the art of the narrative to focus on the experience of using your service/product
• Anexcellentwayofvisualizingthebusiness process and understanding the impact on the customer’s journey
• Aneffectivewaytoprototypeyourserviceconcept
Prototyping
• Afterhavingdevelopedpersonaandstoryboard
• Prototypingcanbedevelopedthroughsoftware, or build with physical materials (e.g., Blue-Tak, paper straws, cardboard, Lego) to give a tangible feel for the product or service
• Stakeholderscaninteractwithprototypeand provide feedback on design
• To help participants empathise more deeply with the problem situation
Opportunity Canvas
• To document the use case after having elaborated it through design thinking
• It provides a basis for the preparation of a business case and presentation of the analytics use case to senior management for approval and funding
Big Data and Algorithm Ethics
Why should we care about ethics of business analytics
• Ethics of big data • Ethics of algorithm
Framework for big-data ethics
Framework for big-data ethics (Davis 2012, p.3. Adopted with permission from O’ )
Sourced from joyoftech.com
• Atthebasiclevel,bigdataistransformingthe way people, organizations, and government understand and think about identity (e.g., IP)
• Atadeeperlevel,bigdataandanalyticshasthe power to influence individual’s capability to define their own identify (e.g., options are limited and shaped based on big data profiling)
Reputation
• Organization can use data to form an opinion, thereby limiting an individual’s ability to shape their reputation and who they are perceived to be
• Giventheincreasinguseofbigdataandanalytics in business applications (e.g., jobs, housing, and credit), the effects of damaging someone’s online reputation can have deep and far-reaching consequences
Ownership of personal data
• The guiding principles are quite complex and still in development, firms are facing a variety of ethical questions and dilemmas over data ownership
• Thecomplexofownershipstemsfromthe tension between the belief that people should have control over their personal data and the belief that the collectors of information have in effect ‘created’ the data and thus have control
Ownership of personal data
• The key differentiator between ownership of data and ownership of physical goods
• Makes an analogy that the control and ownership of data is like housing
Spectrum of organizational involvement in the creation of data
• Threeapproachesto collect personal data
• Conflictoverdata use is at its highest when organizations are deriving insights from data
Adopted from the Framework for big-data ethics (Davis 2012)
Circle of ethical decision-point activities
Adopted from Davis (2012)
• It is centered on identifying how the organization’s core values relate to the use of big data
• Ethical inquiry is crucial, given the potential for heterogeneity of opinions on big data within an organization and the differences in stakeholder motivations and external pressures
Sample questions for inquiring into big-data values (Davis 2012)
• To identify whether its interaction with big data can lead to consequences for the four aspects of the framework for big data ethics (i.e., identify, privacy, ownership, and reputation)
• To assess potential impacts the firm has to conduct a thorough analysis of its data-handling practices, focusing on both people and systems
➢ Who has access to data
➢ The trustworthiness of the people being entrusted with that data
➢ What would be the magnitude of the consequences if the firm’s data were leaked or exposed (e.g., by hackers or by a disaffected employee)
Articulation
• Assess whether a firm’s stated values are aligned with its actions
• Firmsisabletoidentifygapsinitsethical practices by comparing the results of the analysis with the results of the inquiry
• Articulation is completed when the firm has established a document detailing where values and actions fail to align
• Steps taken to close the gaps identified through the articulation of value-action alignment
• Actions need to be based on the organization’s core values given the rapid evolution of big data and big-data technologies
Ethics of algorithms
• An algorithm is a sequence of steps that a machine follow to accomplish a task
• Algorithms are necessary for gaining insights from data
• As analyst objectives become more complex and prescriptive, algorithms are increasingly taking on the form of a ‘black box’ – a process where the internal working of transforming the data inputs to outputs are completely unknown (e.g., Facebook news feed)
The influence of algorithm
• Algorithms are ultimately designed by people, who are far from perfect and filled with a variety of biases in judgement
• Algorithms also rely on data, which can be biased, being based on people’s interpretation of what makes good data
• If encoded with biases, the consequences can be far reaching given their wide application in the society (e.g., job matching, loan approval, online dating, and policing)
Managing algorithms
• Beexplicitaboutthealgorithm’sgoal
• Choose the best data inputs
• Understandthelimitationsofanalgorithm
Adopted from Luca et al. (2016)
Ethical concerns in social media data
• Privateversuspublic • Informedconsent
• Anonymity
• Riskofharm
Decision flow chart for publication of Tweets
(Adopted from Williams et al., 2017)
Regulation
• A significant and impactful development in the data regulation arena is the European Union’s General Data Protection Regulation (GDPR) initiative
• It regulates how organizations collect, store, and process personally identifiable information about EU citizens
• It further provides individuals with rights to explanations – a meaningful explanation of the logic involved in automated decision-making systems, such as customer profiling
Questions?
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com