task models what people do what things they work with what they - PDF document

What is Task Analysis? chapter 15 Methods to analyse people's jobs: task models – what people do – what things they work with – what they must know An Example Approaches to task analysis • Task decom position • in order to clean the house – splitting task into (ordered) subtasks • get the vacuum cleaner out • fix the appropriate attachments • Knowledge based techniques • clean the rooms – what the user knows about the task and how it is organised • when the dust bag gets full, empty it • put the vacuum cleaner and tools away • Entity/ object based analysis – relationships between objects, actions and the people who perform them • m ust know about: • vacuum cleaners, their attachments, dust bags, • lots of different notations/ techniques cupboards, rooms etc. Differences from other general method techniques • observe System s analysis vs. Task analysis system design - focus - t he user • collect unstructured lists of words and actions Cognitive m odels vs. Task analysis • organize using notation or diagrams internal mental state - focus - external actions practiced ` unit' task - focus - whole job 1

Task Decomposition Textual HTA description Hierarchy description ... Aims: describe the actions people do 0. in order to clean the house 1. get the vacuum cleaner out structure them within task subtask hierarchy 2. get the appropriate attachment describe order of subtasks 3. clean the rooms 3.1. clean the hall 3.2. clean the living rooms Variants: 3.3. clean the bedrooms Hierarchical Task Analysis (HTA) 4. empty the dust bag 5. put vacuum cleaner and attachments away m ost common CTT (CNUCE, Pisa) ... and plans uses LOTOS temporal operators Plan 0: do 1 - 2 - 3 - 5 in that order. when the dust bag gets full do 4 Plan 3: do any of 3.1, 3.2 or 3.3 in any order depending on which rooms need cleaning N.B. only the plans denote order Generating the hierarchy Tasks as explanation 1 get list of tasks • imagine asking the user the question: what are you doing now? 2 group tasks into higher level tasks • for the same action the answer may be: 3 decompose lowest level tasks further t yping ctrl-B m aking a word bold Stopping rules em phasising a word How do we know when to stop? editing a docum ent I s “ em pty the dust bag” sim ple enough? w riting a letter Purpose: expand only relevant tasks preparing a legal case Motor actions: lowest sensible level HTA as grammar parse scenario using HTA get out cleaner 1. • can parse sentence into letters, nouns, noun fix carpet head 2. phrase, etc. clean dinning room 3.2. 3. clean m ain bedroom 0. 3.3. em pty dustbag 4. noun phrase clean sitting room 3.2. syntax put cleaner away 5. det noun 0. in order to clean the house 1. get the vacuum cleaner out . . . . . . . . . . . . 2. get the appropriate attachment letter lexical 3. clean the rooms 3.1. clean the hall 3.2. clean the living rooms The cat sat on the mat. 3.3. clean the bedrooms 4. empty the dust bag 5. put vacuum cleaner and attachments away 2

Diagrammatic HTA Refining the description Given initial HTA (textual or diagram ) How to check / improve it? Some heuristics: paired actions e.g., where is ` turn on gas' restructure e.g., generate task ` make pot' balance e.g., is ` pour tea' simpler than making pot? generalise e.g., make one cup … .. or more Refined HTA for making tea Types of plan fixed sequence - 1.1 then 1.2 then 1.3 optional tasks - if the pot is full 2 wait for events - when kettle boils 1.4 cycles - do 5.1 5.2 while there are still empty cups tim e-sharing - do 1; at the same time ... discretionary - do any of 3.1, 3.2 or 3.3 in any order m ixtures - most plans involve several of the above waiting … Knowledge Based Analyses • is waiting part of a plan? … or a task? Focus on: • generally Objects – used in task – task – if ‘busy’ wait Actions – performed • you are actively waiting – plan – if end of delay is the event • e.g. “ when alarm rings” , “ when reply arrives” + Taxonomies – • in this exam ple … represent levels of abstraction – perhaps a little redundant … – TA not an exact science see chapter 19 for more on delays! 3

Knowledge–Based Example … Task Description Hierarchy motor controls Three types of branch point in taxonom y: steering steering wheel, indicators XOR – norm al taxonom y engine/speed object in one and only one branch direct ignition, accelerator, foot brake gearing clutch, gear stick AND – object m ust be in both lights m ultiple classifications external headlights, hazard lights OR – weakest case internal courtesy light can be in one, m any or none wash/wipe wipers front wipers, rear wipers wash/wipe AND washers front washers, rear washers function XOR heating temperature control, air direction, wipe front wipers, rear wipers fan, rear screen heater wash front washers, rear washers parking hand brake, door lock position XOR radio numerous! front front wipers, front washers rear rear wipers, rear washers Larger TDH example More on TDH kitchen item AND Uniqueness rule: /____shape XOR – can the diagram distinguish all objects? / |____dished mixing bowl, casserole, saucepan, / | soup bowl, glass e.g., plate is: / |____flat plate, chopping board, frying pan kitchen item/shape(flat)/function{preparation,dining(for food)}/ /____function OR {____preparation mixing bowl, plate, chopping board nothing else fits this description {____cooking frying pan, casserole, saucepan {____dining XOR Actions have taxonom y too: |____for food plate, soup bowl, casserole kitchen job OR |____for drink glass |____ preparation beating, mixing |____ cooking frying, boiling, baking N.B. ‘ /|{ ’ used for branch types. |____ dining pouring, eating, drinking Abstraction and cuts Entity-Relationship Techniques After producing detailed taxonom y Focus on objects, actions and their relationships ‘cut’ to yield abstract view Sim ilar to OO analysis, but … That is, ignore lower level nodes – includes non-com puter entities e.g. cutting above shape and below dining, plate becomes: – em phasises dom ain understanding not im plem entation kitchen item/function{preparation,dining}/ Running exam ple This is a term in Knowledge Representation Gram m ar ‘Vera's Veggies’ – a market gardening firm (KRG) owner/ manager: Vera Bradshaw These can be m ore com plex: employees: Sam Gummage and Tony Peagreen various tools including a tractor ` Fergie‘ e.g. ‘beating in a mixing bowl’ becomes: kitchen job(preparation) using a two fields and a glasshouse kitchen item/function{preparation}/ new computer controlled irrigation system 4

Objects Attributes Start with list of objects and classify them : To the objects add attributes: Concrete objects: Object Pum p3 sim ple – irrigation pum p sim ple things: spade, plough, glasshouse Attributes : status: on/ off/ faulty Actors: capacity: 100 litres/ m inute hum an actors : Vera, Sam, Tony, the customers what about the irrigation controller? Com posite objects: sets : t he team = Vera, Sam, Tony N.B. need not be computationally complete tuples : tractor may be < Fergie, plough > Actions Actions (ctd) List actions and associate with each: im plicit agents – read behind the words ` the field was ploughed' – by whom ? agent – who perform s the actions patient – which is changed by the action indirect agency – t he real agent? instrument – used to perform action ` Vera program m ed the controller to irrigate the field' m essages – a special sort of action exam ples: ` Vera told Sam to ... ' Sam ( agent ) planted ( action ) the leeks ( patient ) Tony dug the field with the spade ( instrum ent ) rôles – an agent acts in several rôles Vera as w orker or as m anager example – objects and actions Events Object Sam human actor Object glasshouse simple … when something happens Actions : Attribute : S1: drive tractor humidity: 0-100% • performance of action S2: dig the carrots Object Irrigation Controller ‘Sam dug the carrots’ non-human actor Object Vera human actor – the proprietor Actions : • spontaneous events Actions : as worker IC1: turn on Pump1 V1: plant marrow seed IC2: turn on Pump2 ‘the m arrow seed germ inated’ V2: program irrigation controller IC3: turn on Pump3 ‘the hum idity drops below 25% ’ Actions : as manager V3: tell Sam to dig the carrots Object Marrow simple • timed events Actions : Object the men composite M1: germinate ‘at m idnight the controller turns on’ Comprises : Sam, Tony M2: grow 5