Intro Info Visualization
Intro Info Visualization CS 4460
Popular in Course
Popular in ComputerScienence
This 0 page Class Notes was uploaded by Alayna Veum on Monday November 2, 2015. The Class Notes belongs to CS 4460 at Georgia Institute of Technology - Main Campus taught by John Stasko in Fall. Since its upload, it has received 15 views. For similar materials see /class/234086/cs-4460-georgia-institute-of-technology-main-campus in ComputerScienence at Georgia Institute of Technology - Main Campus.
Reviews for Intro Info Visualization
Report this Material
What is Karma?
Karma is the currency of StudySoup.
Date Created: 11/02/15
Cognitive Issues amp User Tasks if 2 72 CS 44607450 Information Visualization January 22 2009 John Stasko Outline Overview 1 Role How visualizations aid cognition 2 Tasks What does the visualization assist Spring 2009 CS 44607450 u Basic Premise a Understanding the cognitive aspects is the crucial part of InfoVis Visualization is simply a tool useful for aiding analysis exploration comprehension and understanding Discussed the role of external cognition aids briefly earlier in intro more now Spring 2009 CS 44607450 3 How Are Graphics Used 5 l r What does a visualization or graphic image provide for us Spring 2009 CS 44607450 4 How Are Graphics Used h Larkin amp Simon 87 investigated usefulness of graphical displays Graphical visualization could support more efficient task performance by Allowing substitution of rapid perceptual influences for difficult logical inferences Reducing search for information required for task completion 0 Sometimes text is better however Spring 2009 CS 44607450 5 Norman s Thoughts Cognitive Artifacts Wonderful discussion on p 49 Matching Representation to Task Tictactoe flight schedules 0 Representations Aid Info Access and Computation Medical prescriptions Roman numerals maps amp legends Naturalness and Experiential Cognition Visual Representations Chapter 3 from Sprmg 2009 cs 44607450 Things That Make Us Smart 6 Revisit Visualization u Often thought of as process of making a graphic or an image Really is a cognitive process Form a mental image of something Internalize an understanding 0 The purpose of visualization is insight not pictures Insight discovery decision making explanation Spring 2009 CS 44607450 7 Revisit Mam Idea Visuals help us think Provide a frame of reference a temporary storage area Cognition gt Perception Pattern matching 0 External cognition aid Role of external world in thinking and reason Larkin amp Simon 87 Card Mackinlay Shneiderman 98 Spring 2009 CS 44607450 8 Revisit Visualization Definition The use of computersupported interactive visual representations of data to amplify cognitionquot From Card Mackinlay Shneiderman 98 Spring 2009 CS 44607450 9 Examine More Closely 5 r r o What does amplify cognition mean Discuss Spring 2009 CS 44607450 10 Another View Leverage Hutchins theory of distributed cognition DCog to explain the value and utility of infovis Use DCog as a supporting theoretical framework for infovis Liu Nersessian Stasko TVCG 08 Spring 2009 CS 44607450 11 Amplifying Cognition jg Hutchins argues that tools don t amplify or scaffold cognition a more traditional cognitive science view Eg Our memory isn t amplified Instead tools help transform the analytic process into another more doable one Spring 2009 CS 44607450 12 u Distributed Cognition a Cognitive system is composed of people and the artifacts they use Cognition isn t only internal Changes in external representation spur changes in internal representation and understanding 0 It is interaction with the external representations that drives this process Spring 2009 CS 44607450 13 More Details ta OK so now let s talk about the analytic process in more detail and specifically how visualization can play a role Spring 2009 CS 44607450 14 u Understanding a 0 People utilize an mentalinternal model that is generated based on what is observed B Tversky calls the internal model a cognitive map Think about that term Spring 2009 CS 44607450 Example 2quot You re taking the MARTA train to get to Georgia State University A You have some existing internal model of the system stops how to get there On train you glance at MARTA map for help Refines your internal model clarifying items and extending it Note that it s still not perfect no internal model ever is Spring 2009 CS 44607450 Cognitive Map Just don t have one big one 0 Have large number of these for all different kinds of things Collection of cognitive maps gt Cognitive colage Spring 2009 CS 44607450 17 1 Process Models 2w Recall the user and cognitive models from HCI Process by which a person looks at a graphic and makes some use of it A number of substeps probably exist Can you describe process Spring 2009 CS 44607450 18 Process Model 1 Robert Spence Naygaton Creation and interpretation of an internal mental model Spring 2009 CS 44607450 19 Navigation Spring 2009 CS 44607450 20 Interpretation Can someone explain that Spring 2009 CS 44607450 21 Interpretation 2a Content is the display on screen Modeling of that pattern results in cognitive map Interpretation ah variables x and y are related leads to new view that generates an idea for a new browsing strategy Look at the display again with that Spring 2009 CS 44607450 22 Process Model 2 Card Mackinlay Shneiderman book 0 Knowledge crystallization task Gather info for some purpose make sense of it by constructing a representational framework and package it into a form for communication or action Spring 2009 CS 44607450 23 Knowledge Crystallization Information foraging Search for schema representation Instantiate schema Problem solve to trade off features Search for a new schema that reduces problem to a simple tradeoff Package the patterns found in some output product From CMS 98 Spring 2009 CS 44607450 24 How Vis Amplifies Cognition Increasing memory and processing resources available Reducing search for information Enhancing the recognition of patterns Enabling perceptual inference operations 0 Using perceptual attention mechanisms for monitoring Encoding info in a manipulable medium Spring 2009 cs 44607450 25 P rocess o 0 0 Raw Data Visual Views data tables Structures w Data Visual View transformations mappings transformations Spring 2009 CS 44607450 26 Knowledge Crystallization Overview Filter Detailsondemand gmws Extra ct ea rc ue 5 r 1 r gt Compose Reorder Cluster 39 Average mote Detect pattern Abstra ct Spring 2009 cs 44607450 were 27 2 User Tasks What things will people want to accomplish using information visualizations Earlier we briefly discussed search vs browsing Spring 2009 CS 44607450 Browsing vs Search Important difference in activities Appears that information visualization may have more to offer to browsing Butbrowsing is a softer fuzzier activity 0 So how do we articulate utility Maybe describe when it s useful When is browsing useful Spring 2009 CS 44607450 29 Browsing 39 g g Useful when Good underlying structure so that items close to one another can be inferred to be similar Users are unfamiliar with collection contents Users have limited understanding of how system is organized and prefer less cognitively loaded method of exploration Users have difficulty verbalizing underlying information need Information is easier to recognize than describe Lin 97 Spring 2009 CS 44607450 30 Thought Maybe infovis isn t about answering questions or solving problems hmmm Maybe it s about asking better questions Spring 2009 CS 44607450 31 Tasks MW a OK but browsing and search are very high eve Let s be more specific Spring 2009 CS 44607450 32 Example from Earlier Which cereal has the mostleast potassium Questions Is there a relationship between potassium and fiber so are there any outliers Which manufacturer makes the healthiest cereals 22 P 2 as m m an m o 2 so 1 m 3i WWW c a 55 3 2 m A u s u 3 4 N A rm 4 m N 7 an E a iBE r v m in quotmm a u 3 y a M u Enema r 2 ms i i 45 u Luca m c c 55 o 7 ma 0 a m N n N lt 4 im N w gt1 N i c N 4 HE E 1 il N n m a i a i i An N i 1 K i 55 N i y N 3 m 5 m r u n 1 g m r r m n c 25 v rlt v r N 54 WW 3s v 2 im 1 r i a m Whams may m s i In Spring 2009 CS 44607450 33 Exercise o What are the types of tasks being done here Can you think of others 7 Let s develop a list Spring 2009 CS 44607450 34 Task Taxonomies quot f c Number of different ones exist important to understand what process they focus on Creating an artifact Human tasks Tasks using visualization system Spring 2009 CS 44607450 35 User Tasks 2g Wehrend amp Lewis created a lowlevel domain independent taxonomy of user tasks in visualization environments Eleven basic actions identify locate distinguish categorize cluster distribution rank compare within relations compare between relations associate correlate Vis 90 Spring 2009 cs 44607450 36 Another Perspective 1 Shneiderman proposed task x data type taxonomy to understand what people do with visualization Mantra Overview first zoom and filter then details on demand Design paradigm for infovis systems Shneiderman VL 96 Spring 2009 CS 44607450 37 Taxonomy 39 g Data Types Tasks 1 1D 1 Overview 2 2D 2 Zoom 3 3D 3 Filter 4 Temporal 4 Detailsondemand 5 ND 5 Relate 6 Tree 6 History 7 Network 7 Extract Spring 2009 CS 44607450 38 Another Task Taxonomy Amar Eagan amp Stasko InfoVis 05 0 Discuss Spring 2009 CS 44607450 39 Background Use commercial tools class assignment coming soon Students generate questions to be answered using commercial infovis systems Data sets Domain Data Attributes Questions cases Generated Cereals 78 15 107 Mutual funds 987 14 41 Cars 407 10 153 Filn39s 1742 10 169 Gmoery surveys 5164 8 126 Generated 596 total analysis tasks Spring 2009 CS 44607450 40 Terminology g Data 6556 An entity in the data set Attribute A value measured for all data cases 39 Aggrega on function A function that creates a numeric representation for a set of data cases eg average count sum Spring 2009 CS 44607450 44 1 Retrieve Value General Description Given a set of specific cases nd attributes of those cases Examples What is the mileage per gallon of the Audi 39l39l39 How long is the movie Gone with the Wind Spring 2009 CS 44607450 45 2 Filter General Description Given some concrete conditions on attribute values nd data cases satisfying those conditions Examples What Kellogg39s cereals have high fiber What comedies have won awards Which funds underperformed the SPSOO Spring 2009 CS 44607450 46 3 Compute Derived Value General Description Given a set of data cases compute an aggregate numeric representation of those data cases Examples What is the gross income of all stores combined How many manufacturers of cars are there What is the average calorie content of Post cereals Spring 2009 CS 44607450 4 Find Extremum N General Description Find data cases possessing an extreme value of an attribute over its range within the data set Examples What is the car with the highest MPG What directorfilm has won the most awards What Robin Williams lm has the most recent release date Spring 2009 CS 44607450 5 Sort General Description Given a set of data cases rank them according to some ordinal metric Examples Order the cars by weight Rank the cereals by calories Spring 2009 CS 44607450 49 6 Determine Range V General Description Given a set of data cases and an attribute of interest find the span of values within the set Examples What is the range of film lengths What is the range of car horsepowers What actresses are in the data set Spring 2009 CS 44607450 50 7 Characterize Distribution General Description Given a set of data cases and a quantitative attribute of interest characterize the distribution of that attribute s values over the set Examples What is the distribution of carbohydrates in cereals What is the age distribution of shoppers Spring 2009 CS 44607450 51 8 Find Anomalies N General Description Identify any anomalies within a given set of data cases with respect to a given relationship or expectation eg statistical outliers Examples Are there any outliers in protein Are there exceptions to the relationship between horsepower and acceleration Spring 2009 CS 44607450 52 9 Cluster General Description Given a set of data cases find clusters of similar attribute values Examples Are there groups of cereals w similar fatcaloriessugar Is there a cluster of typical film lengths Spring 2009 CS 44607450 53 1 0 Correlate a General Description Given a set of data cases and two attributes determine useful relationships between the values of those attributes Examples Is there a correlation between carbohydrates and fat Is there a correlation between country of origin and MPG Do different genders have a preferred payment method Is there a trend of increasing film length over the years Spring 2009 CS 44607450 54 DiscussionReflection Compound tasks Sort the cereal manufacturers by average fat contentquot Compute derived value Sort Which actors have costarred with Julia Robertsquot Filter Retrieve value Spring 2009 CS 44607450 DiscussionReflection What questions were left out Basic math Which cereal has more sugar Cheerios or Special K Compare the average MP of American and Japanese cars Uncertain criteria Does cereal X Y Z s d tasty What are the characteristics of the most valued customers Higherlevel tasks How do mutual funds get rated Are there car aspects that Toyota has concentrated on More qualitative comparison How does the Toyota RAV4 compare to the Honda CRV What other cereals are most similar to Trix Spring 2009 CS 44607450 u Concerns InfoVis tools may have influenced students questions Graduate students as group being studied How about professional analysts Subjective Not an exact science Data was really quantitative so may get a different set of tasks for relationalgraph data A See Lee et al BEle 06 Spring 2009 CS 44607450 57 Contributions lg Set of grounded lowlevel analysis tasks Potential use of tasks as a languagevocabulary for comparing and evaluating infovis systems Spring 2009 CS 44607450 58 4 Can We Be More Is InfoVis helping people enough 0 What do we need to do to provide even more value Spring 2009 CS 44607450 59 Providing Better Analysis N r Combine computational analysis approaches such as data mining with infovis Too often viewed as competitors in past 0 Each has something to contribute S h n e i d e rm a n Infomatbn l39sua39zatbn 02 Spring 2009 cs 44607450 60 Issues 1 Issues influencing the design of discovery tools Statistical Algorithms vs Visual data presentation Hypothesis testing vs exploratory data analysis Pro s and Con s Spring 2009 CS 44607450 61 I ng ewamp o Hypothesis testing Advocates By stating hypotheses up front limit variables and sharpens thinking more precise measurement Critics Too far from reality initial hypotheses bias toward finding evidence to support it o Exploratory Data Analysis Advocates Find the interesting things this way we now have computational capabilities to do them Skeptics Not generalizable everything is a special case detecting statistical replationships does not infer cause and effect Spring 2009 CS 44607450 62 u Recommendations Integrate data mining and information visualization Allow users to specify what they are seeking Recognize that users are situated in a social context 0 Respect human responsibility Spring 2009 CS 44607450 63 Another Question 2w Are the visualizations helping with exploratory analysis enough Are they attempting to accomplish the right goals Spring 2009 CS 44607450 64 Status Quo Limitations Current Information Visualization systems inadequately support decision making Limited Affordances Predetermined Representations Decline of Determinism in DecisionMaking Representational primacy versus Analytic primacy Amar amp Stasko TlCG 05 Spring 2009 cs 44607450 55 HighLevel Tasks P Complex decisionmaking especially under uncertainty 0 Learning a domain Identifying the nature of trends Predicting the future Spring 2009 CS 44607450 66 Analytic Gaps 0 Analytic gaps obstacles faced by visualizations in facilitating higherlevel analytic tasks such as decision making and learningll Worldview Gap Rationale Gap Worldview Ga Rationale Ga Representation of Data Dilisrdata Con dence m data Otherviews7 Con dence m rela on smps Idenlify wew elements Perceiving Useful Relationships Analyst Perceptual EXplaining Relationships Spring 2009 cs 44607450 67 Knowledge Precepts For narrowing these gaps v WorldviewBased Precepts Did we show the right thing to the userquot Determine Domain Parameters Expose Multivariate Explanation Facilitate Hypothesis Testing RationaleBased Precepts Will the user believe what they seequot Expose Uncertainty Concretize Relationships Expose Cause and Effect Spring 2009 cs 44607450 68 Application of Precepts Fly 2 Emu bars which we have added in led would be a shnple way m Increase cmlmmca m ha ds Isa 01 dmavence nemasn lwo aggraga 7 Hons F39mure Bken mm the Seen syslem by V smle Declsbns lm Lg 3 Ms lhemescape vanamn a ows umumams Mm rmssmg maladala shown as mu m Ihe uppev mack vegvon m parmxpale m anaws such as Ihe relerence relalmrmhip sham chlure emuva Mamas Dlakoumms Spring 2009 CS 44607450 69 Application of Precepts F 4 The mew quot EDGE Pm lm w m qu39ck y exam Fug 51NSP FIE uses mmmta schUnQ o nawgammna shcas arussr P S quot39E mes 39Cn39m39a mr 39quot e39 quota quota m39 dahnad cumem groups F mure pmducau m and mum mm pamxssnon a Pacl c Nonnwasx Nahunal Labuamw winch Is managed and mayaled w ma Banana Mammal nsmua an hanau al lha us Dapanmem or Energy Spring 2009 CS 44607450 70 Put Them Together Combine the ideas 7 Use computational statistical analysis more 7 Cater to the user s analytic reasoning needs 0 And put together with infovis o Leads to Spring 2009 CS 44607450 71 Visual Analytics o The science of analytical reasoning facilitated by interactive visual interfaces Combines 7 Data analysis 7 Infovis 7 Analytical reasoning Grew from view that infovis was neglecting these other aspects 7 True Thomas amp Cook 939 Illumha hg the Path Spring 2009 cs 44607450 72 Visual Analytics h Grew from stimulus in the homeland security area Need for better data analysis methods Really big data Topic for entire day later in term Spring 2009 CS 44607450 73 HW2 Due Tuesday Questions Spring 2009 CS 44607450 74