CMPUT 301: Lecture 16 Task Analysis II Lecturer: Martin Jagersand Department of Computing Science University of Alberta Notes based on previous courses.

CMPUT 301: Lecture 16Task Analysis II

Lecturer: Martin JagersandDepartment of Computing Science

University of Alberta

Notes based on previous courses byKen Wong, Eleni Stroulia

Zach Dodds, Martin Jagersand

2

Overview:

• Last time:– Hierarchical analysis– Temporal plans– Knowledge based analysis

• Today:– Entity-relation based analysis– Using several of the above together

3

Entity-Relation-Based Techniques

• Understanding objects and actions:– list all objects involved– focus on the relationships between objects and

actions, rather than similarities– link actions with objects which help perform them– “object-based”

4

Entity-Relation-Based Techniques

• Objects of gardening firm:– tool shed with spade, garden fork, hoe– Ferguson tractor, with plow and harrow– owner (Vera)– two employees (Sam and Tony)– two growing fields– large glasshouse with humidity sensor– computer-controlled irrigation system

team

The kit

5

Entity-Relation-Based Techniques

• Classify objects into three types:– simple objects

– “things”

– e.g., spade, plow, glasshouse

– actors– humans and non-humans that do things

– e.g., Vera, Sam, Tony, Fergie, computer

– composite objects– objects which consist of other objects

– e.g., team, men, kit

6

Entity-Relation-Based Techniques

• Look at some attributes of objects:– Object Pump3 simple — irrigation pump

– Attributes– status: on/off/faulty

– capacity: 100 litres/minute

7

Entity-Relation-Based Techniques

• Only enough detail about attributes to describe their use in tasks:– Actions are e.g.,

– turning on and off irrigation pumps, done by user actor

– may not need pump capacity – only needed by repairman for replacement

• We can drop unwanted details later.

8

Entity-Relation-Based Techniques

• Look at actions that change the state of something:– agent (often an actor), action, patient– associated attributes– e.g.,

– Sam planted carrots

– gardener dug soil with spade

– irrigation controller turning on water

9

Entity-Relation-Based Techniques

• Message actions:– communication– perhaps in prelude to another action– e.g.,

– Vera telling Sam to dig the carrots

10

Entity-Relation-Based Techniques

• Identify roles:– actors fill in specific roles– an actor can have multiple roles– a role can have multiple actors– e.g.,

– Vera as worker

– Vera as manager

– Sam as worker

11

Entity-Relation-Based Techniques

• Object Sam human actor– Actions

– s1: drive tractor– s2: dig the carrots

• Object Vera human actor– Actions as worker

– v1: plant seed– v2: program irrigation controller

– Actions as manager– v3: tell Sam to dig the carrots

12

Entity-Relation-Based Techniques

• Object the men composite– Comprises

– {Sam, Tony}

• Object glasshouse simple– Attribute

– humidity: 0–100%

13

Entity-Relation-Based Techniques

• Object Irrigation Controller non-human actor– Actions

– ic1: turn on Pump1

– ic2: turn on Pump2

– ic3: turn on Pump3

14

Entity-Relation-Based Techniques

• Identify events:– things that happen– Different types e.g.,

– germination of “seed”(spontaneous action of seed, non-actor)

– humidity drops below 25%(no associated object)

– at midnight(timed)

15

Entity-Relation-Based Techniques

• Identify relationships between objects, actions, and events:– object to object

– e.g., Sam is subordinate to Vera

– action to object– e.g., (Sam) digs the carrots

16

Entity-Relation-Based Techniques

• Relations object–object– located in: ( Pump3, glasshouse )– located in: ( Pump1, Field1 )

17

Entity-Relation-Based Techniques

• Relations action–object– patient ( v3, Sam )

– i.e., Vera tells Sam to dig

– patient ( s2, the carrots )– i.e., Sam digs the carrots

– instrument ( s2, spade )– i.e., Sam digs the carrots with the spade

18

Entity-Relation-Based Techniques

• Relations action–event– before ( v1, m1 )

– i.e., seed must be sown before it can germinate

– triggers ( ev1, ic3 )– i.e., when humidity drops below 25%, the controller

turns on pump 3

– causes ( v2, ic1 )– i.e., the controller turns on pump 1 because Vera

programmed it

– etc.

19

Combining Approaches

• 0 To grow carrots– 1 Vera sows the seed– 2 seed germinates– 3 Vera programs controller– 4 controller waters field– 5 carrots grow– 6 Sam hoes– 7 Tony harvests carrots

20

Combining Approaches

• Plans:– plan 0

– do 1, 2, 5, when crop is mature do 7

– when rainfall is low, do 3, 4

– when weeds grow, do 6

21

Combining Approaches

• Look at a particular object and all the actions in which it participates:– life cycle diagram– many tasks

Combining Approaches

22

Combining Approaches

23

Combining Approaches

• Use knowledge-based (taxonomic) analysis:– identify commonalities of actions and relations– build inheritance hierarchy– “object-oriented”– e.g., Sam and Tony could do some things in

common

24

Sources of Information

• Need to collect information about tasks from various sources:– documentation– observation– initial analysis– sorting and classification

25

Sources of Task Information

• Documentation:– e.g., existing manuals, training materials, job

descriptions– job descriptions may only describe what people

are supposed to do, not what they actually do– equipment manuals may only describe

functions or features, not tasks

26

Sources of Task Information

• Observation:– formal or informal– in the field on in the lab– passive or active– note taking– video taping– talk aloud– walkthroughs– confirm task analyses

27

Sources of Task Information

• Interviews:– pose questions to domain experts directly– follow up on responses with more questions,

for more detail

28

Sources of Task Information

• Initial analysis:– following on from manuals, observation, or

interviews– build lists of objects (nouns) and actions

(verbs)

29

Sources of Task Information

• Sorting and classification:– use note cards of task objects– have a domain expert sort cards into piles by

similarity– use spatial arrangement– compare groupings among experts– use an outliner

30

Uses of Task Analysis

• Manuals:– use hierarchical task analysis to structure

training materials(“how to do it”)

– use knowledge-based techniques to structure a course or textbook(concepts and foundations)

31

Uses of Task Analysis

• Requirements capture:– task analysis of existing procedures– help in eliciting what the new system must do

and guiding its design– how the system is used versus what the system

does– compare new and old procedures

32

Uses of Task Analysis

• Detailed interface design:– use task decomposition trees to design menus

and layout dialogs– or use roles and tasks within each role to design

menus– or use objects and what can be done to it to

design menus– use taxonomies for grouping– avoid too much “movement”

33

End

• What did I learn today?

• What questions do I still have?