Explosion of Data CSE 564 Visualization Visual Analytics 82 000 fingerprints are matched every day against INS database with 40 million records 100 million VISA credit card transactions per day 300 million phone long distance calls on ATT s network per day 7 million IP packets per second on DE CIX backbone Klaus Mueller there is NO chance to visualize all these data Computer Science Department Stony Brook University Problems With Scalability Must be scalable to number of data points number of dimensions data sources diversity of data sources number of users diversity of users and tasks quality of the data The Goal of Visualization Ease understanding of the data by providing an effective visual representation Amplify Perception Visual Analytics comes to the rescue Detect the Expected Discover the Unexpected What is Visual Analytics Visualization plus interaction HCI data processing analytics story telling scientific approach Agenda setting book http nvac pnl gov agenda stm Visual Analytics Method The Daniel Keim Mantra of Visual Analytics Analyze First Show the Important Zoom Filter and Analyze Further Details on Demand but also Interaction intelligent computing AI machine learning behavioral psychology cognitive science human factors The triangle of Visual Analytics VA Visual Analytics is the science of analytical reasoning supported by a highly interactive visual interface VA Analytics Intelligence Analysis Intelligence analysis is challenging Huge amounts of data Low signal vs noise SNR Many data types text images video sensor data etc Uncertainty Visualization Use of Visualization Visual perception high bandwidth fast screening of a lot of data pattern recognition higher level cognition Interaction direct manipulation two way communication Contradictions Omissions Recall intro lecture on the human visual system Use of Visualization Visual perception high bandwidth fast screening of a lot of data pattern recognition higher level cognition Interaction direct manipulation two way communication Focus vs Periphery Humans tend to overlook ignore non focus and unexpected objects even when very close and obvious note the Visual Analytics slogan Detect the Unexpected Humans also have limited working memory fine details are quickly forgotten when focus changes big effect in animated or interactive visualizations need to preserve temporal context Recall intro lecture on the human visual system But humans are imperfect Change Blindness Thoroughly studied by Dan Simons U Illinois see http www dansimons com index html Videos Change Blindness Thoroughly studied by Dan Simons U Illinois see http www dansimons com index html Videos Visual Analytics tools help human analysts cope with insufficient memory visualizations externalize memory allow humans to perform visual queries see C Ware book help human analysts deal with change blindness analytics can detect changes visualization can highlight emphasize these changes Persistence of Mindset Priming Persistence of Mindset Priming Another deficiency of humans humans tend to stick with an opinion for a long time Another deficiency of humans humans tend to stick with an opinion for a long time man woman Young old woman Human Limitations The Magic Number Seven 2 the number of things most people can keep in working memory at one time causes problems for complicated analysis Strategies for Dealing with Complexity Decomposition decompose a complex problem into simpler problems get one s thinking straight in these simpler problems Externalization get the decomposed problem out of one s head and down on paper or on a computer screen in some simplified form shows the main variables parameters or elements of the Excellent book the next slides follow it problem and how they relate to each other Strategies for Dealing with Complexity Decomposition decompose a complex problem into simpler problems get one s thinking straight in these simpler problems Externalization get the decomposed problem out of one s head and down on paper or on a computer screen in some simplified form shows the main variables parameters or elements of the problem and how they relate to each other Recall principles of information visualization and visual analytics overview and detail focus and context analyze filter zoom Multiattribute Utility Analysis You want to choose the best car among various cars What is the best car Lowest maintenance cost Highest resale value Slickest styling Best gas mileage Largest trunk space How to make a decision Car purchase matrix 200 Years Ago Benjamin Franklin s Letter Mentioned his method of solving decision problems Why is the decision problem so difficult folks cannot keep all pros and cons in mind at the same time Solution write down all the pros and cons onto paper in some visible shorthand form allows you make a global judgment effectively Visual analytics many 1 List the important attributes you want to maximize 2 Quantify the relative importance of each attributes 3 Identify the cars you are considering and judge each one ranks on each of attributes 4 Multiply the percentage value by the value of each cars More Formally Problem Description When working on difficult intelligence issues which is the correct explanation which is the most likely outcome Alternative hypothesis 24 Analysis of Competing Hypotheses ACH Used to aid judgment on important issues minimize cognitive limitations Eight Step of ACH 1 Identify hypothesis 2 List evidence Check evidence source Basic insights from cognitive psychology decision analysis scientific method 3 Prepare matrix 5 Draw conclusions 4 Refine matrix 6 Analyze conclusions 7 Report conclusions 8 Identify milestones 25 Step 1 Identify Hypothesis 26 Step 2 List Evidence Hypothesis generation vs hypothesis evaluation generation bring together all possibilities evaluation focus on Don t limit to the evidences current available Disproved vs unproven for a disproved hypothesis there is positive evidence that it is wrong for an unproven hypothesis there is no evidence that it is correct Absence and presence of evidence for example If the dog barked in the night For each hypothesis list supporting and contradicting factors 27 no nobody heard it barked absence 28 Step 3 Prepare Matrix Question will Iraq Retaliate for US Bombing H1 Iraq will not retaliate H2 It will sponsor some minor terrorist actions H3 Iraq is planning a major terrorist attack perhaps against one or more CIA installations 29 Step 4 Refine Matrix 30