Department of Computer Science University of Toronto Department of Computer Science University of Toronto Starting Points Lecture 3 starting points Last Last Week Week Context Contextfor forRE RE What What isis Engineering Engineering Lifecycle Lifecycle models models Systems Systems Theory Theory Stakeholders Importance of Customer Links Who are the stakeholders Boundaries How do you scope the problem This This Week Week Initiating Initiating an an RE RE process process 2000 2003 Steve Easterbrook 2 2000 2003 Steve Easterbrook Department of Computer Science University of Toronto The four worlds Usage World the environment within which the planned system will operate e g people such as managers clerks customers also business processes such as handling a withdrawal a deposit of foreign currency Uses System World what the system does within its operational environment what information it contains and what functions it performs System World e g system records all transactions in a database reports on transactions for a particular account gives account balance builds Development World the development process team schedule required qualities security performance etc e g system to be delivered in 12 months fully tested to MCDC standard etc Development World Source Adapted from Loucopoulos Karakostas 1995 p73 Subject World e g customers accounts transactions for a bank information system Subject World 2000 2003 Steve Easterbrook Department of Computer Science the subject matter of the information system contracts 3 Finding out about the four worlds Maintains information about Usage World Risk Continuous Risk Management Identifying risks through hazard and fault analysis Interviews Interviews Questionnaires Questionnaires Cognitive Cognitive approaches approaches Ethnography Ethnography Needs information about Feasibility How to conduct a feasibility study How to choose which project to persue Next Next Week Week Elicitation Elicitation Techniques Techniques University of Toronto Goals and Scenarios A useful way to organise initial collection of information Stakeholders Stakeholders and and Boundaries Boundaries Goals Goals and and Scenarios Scenarios Feasibility Feasibility and and Risk Risk 4 2000 2003 Steve Easterbrook 5 1 Department of Computer Science Stakeholders Links with customers Successful projects tend to have more links with customer s Technical authors 20 10 0 will prepare user manuals and other documentation for the new system P1 The project manager wants to complete the project on time within budget with all objectives met the customer whoever it is that pays for the new system 2000 2003 Steve Easterbrook 6 P5 P8 P9 P10 P11 Company C1 C2 C3 C4 C5 C6 7 Source Adapted from Keil and Carmel 1995 p37 University of Toronto Customer Developer Links Department of Computer Science Difficulties of Elicitation Customer Support Line Thin spread of domain knowledge The knowledge might be distributed across many sources Marketing and Sales Involvement It is rarely available in an explicit form I e not written down Feedback from testing There will be conflicts between knowledge from different sources Interviews Type of customer Link P3 2000 2003 Steve Easterbrook Department of Computer Science University of Toronto P2 manuf of elec goods want to make sure we are doing better than the competition medium beverage co Business analysts 30 major hotel chain want to make sure the new system is usable and manageable major airline Training and user support staff 40 a telecoms company want to get the requirements right large comp company Systems analysts 50 s w div of h w co want to build a perfect system or reuse existing code custom s w More Successful Project Less Successful Project 60 office auto devel concerned with the features and functionality of the new system Designers packages 70 prog env developer Users CASE tool developer Example stakeholders Links used as a percentage of all possible links From Keil Carmel CACM May 1995 80 manuf s w devel Identify all the people who must be consulted during information acquisition Look for stakeholders associated with each of the four worlds financial s w devel Stakeholder analysis software tool devel Department of Computer Science University of Toronto UNIX tool supplier University of Toronto People have conflicting goals People have different understandings of the problem E mailing list or bulletin board Requirements Prototyping User Interface Prototyping Tacit knowledge The say do problem People find it hard to describe knowledge they regularly use Focus Groups Descriptions may be inaccurate rationalizations of expert behaviour Trade Shows User Groups Limited Observability The problem owners might be too busy solving it using the existing system Presence of an observer may change the problem On site customer representative Facilitated Workshop e g JAD Observational Study E g the Probe Effect and the Hawthorne Effect Custom Package Usability Lab Survey Bias People may not be free to tell you what you need to know 0 2 4 6 8 10 Political climate organisational factors matter 12 People may not want to tell you what you need to know Number of Projects using this type of link 17 projects total The outcome will affect them so they may try to influence you hidden agendas 2000 2003 Steve Easterbrook 8 2000 2003 Steve Easterbrook 9 2 Department of Computer Science University of Toronto University of Toronto Example Psychological Considerations The problem area 1 cognitive verbal rehearsal of tasks 2 associative reinforcement through repetition verbal mediation disappears 3 autonomous compiled no conscious awareness of performance Why this might be difficult Procedural and declarative are different mechanisms Declarative knowledge becomes procedural with repeated application experts lose awareness of what they know and cannot introspect reliably Experts have little or no introspective access to higher order cognitive processes Implicit knowledge There is no document in which the rules for approving loans are written down Conflicting information Different members of the department have different ideas about what the rules are No spoken language offers the necessary precision Knowledge Engineer and Expert must work together to create a suitable language The loan approval process described to you by the loan approval officers is quite different from your observations of what they actually do Different knowledge representations are good for different things