Agenda CPSC 533C Information Visualization Project Update • Motivation: Exploratory browsing? Exploratory Browsing in • The ideal infovis solution: what should it be? Music Space • Related work: displaying query-based results • Prototypes: my proposed solution Heidi Lam • Dataset and implementation November 17, 2004 • List of ongoing and future work Project Idea Query Taxonomy • How can computer tools/interfaces better Specified Target Uncertain Target support exploratory browsing? Navigation: if a map of the Navigation: if a map of the space space is present is present Specified • What is exploratory browsing? Exploration: if not Exploration: if not Location Redundant encoding (target and Single encoding (location) to location) to evaluate if the target evaluate if the target is found is found Search/find with static Browsing with potentially evaluation dynamic evaluation Uncertain Location (i.e., looking for something (i.e., target is ill-defined, and its defined) properties may change/be refined along the process). Query Taxonomy Query Taxonomy Specified Target Uncertain Target Specified Target Uncertain Target Navigation: if a map of the Navigation: if a map of the space Navigation: if a map of the Navigation: if a map of the space space is present is present space is present is present Specified Specified Exploration: if not Exploration: if not Exploration: if not Exploration: if not Location Location Redundant encoding (target and Single encoding (location) to Redundant encoding (target and Single encoding (location) to location) to evaluate if the target evaluate if the target is found location) to evaluate if the target evaluate if the target is found is found is found Search/find with static Browsing with potentially Search/find with static Browsing with potentially evaluation dynamic evaluation evaluation dynamic evaluation Uncertain Uncertain Location Location (i.e., looking for something (i.e., target is ill-defined, and its (i.e., looking for something (i.e., target is ill-defined, and its defined) properties may change/be refined defined) properties may change/be refined along the process). along the process). 1
Query Taxonomy Query Taxonomy Specified Target Uncertain Target Specified Target Uncertain Target Navigation: if a map of the Navigation: if a map of the space Navigation: if a map of the Navigation: if a map of the space space is present is present space is present is present Specified Specified Exploration: if not Exploration: if not Exploration: if not Exploration: if not Location Location Redundant encoding (target and Single encoding (location) to Redundant encoding (target and Single encoding (location) to location) to evaluate if the target evaluate if the target is found location) to evaluate if the target evaluate if the target is found is found is found Search/find with static Browsing with potentially Search/find with static Browsing with potentially evaluation dynamic evaluation evaluation dynamic evaluation Uncertain Uncertain Location Location (i.e., looking for something (i.e., target is ill-defined, and its (i.e., looking for something (i.e., target is ill-defined, and its defined) properties may change/be refined defined) properties may change/be refined along the process). along the process). Two Scenarios at a Record Store Two Scenarios at a Record Store 1. Looking for Ray Charles’ “Come Rain or The goals of these scenarios are different: Come Shine” – With find/search/navigation: want to find the target – Navigate: Go to “Jazz” � Search under “C” � as quickly as possible Find “Ray Charles” � Search among his albums – With exploratory browsing: getting there is half of – Find/Search: “Do you have Ray Charles’ “Come the fun/work? rain or come shine”? 2. Browsing at the “Classical” section � Came across a Jazzified version of Bach � Go to the “Jazz” section � Ray Charles’ album is on display Project Motivation The Ideal InfoVis Solution • Exploratory browsing is not well-supported by To better support exploratory browsing, the current tools interface should … Specified Target Uncertain Target 1. Provide context: to allow users to interpret the Navigate/Explore query results based on their input terms Navigate/Explore Specified � where am I? what am I looking at? Location File explorer File explorer Web browser Web browser 2. Guide navigation: going from the familiar to the unfamiliar � where did I come from? where should I Find/Search Browse Uncertain go next? Location Internet search engine? File searcher Internet search engine 3. Assist refinement of target: based on retrieved results and query terms � what am I looking for? 2
Related Work: Overview Related Work: Spatial 1. Show relationship • Focus on query-criteria based from a single between keywords— search mechanism InfoCrystal (1993) • 4 approaches: 1. Spatial: retrieved results are clusters into groups based on query terms, and displayed spatially 2. List: retrieved results are displayed as a linear list 3. Temporal: retrieved results in the context of timelines 4. Integrated: multi-view with combinations of the above approaches Related Work: Spatial Related Work: Spatial 1. Show relationship between 1. Show relationship between keywords—InfoCrystal keywords—InfoCrystal (1993) (1993) 2. Show clusters only—Lighthouse 2. Show clusters only—Lighthouse (2000) (2000) 3. 3D terrain— BEAD (1993) Related Work: Spatial Related Work: List 1. Show relationship between • Google keywords—InfoCrystal (1993) • Stuff I've Seen (2003) 2. Show clusters only— Lighthouse (2000) 3. 3D terrain—BEAD (1993) 4. Arrange multiple query results spatially —Sparkler (2001) 3
Related Work: Temporal Related Work: Integrated • Milestones in Time • InfoSpace (2003): spatial + temporal (2003) � provides personal events as landmarks on the time line for the retrieved results Prototype I Prototype I • Arranges query • Uses a number of visualization techniques results as a Venn to convey these diagram relationships – to put results in - Colour-coding the search word with primary context of query colours, and the cross- terms area with a mix of those colours – to relate neigbouring regions by a query - Perceptual Layering to term indicate the relative importance of each result region Prototype I Prototype II • Difficult to pack • Uses rectangular squares into non- containers rectangular – Harder to see “Venn” containers relationships, but still relates neighbours � Limits max display with a single query capacity term • Can further cluster • “Piles” music by music record composer (or artist, displays genre, style) 4
Prototype II Prototype II • Semantic zooming • Semantic zooming 1. Full display 1. Full display (composer, title, (composer, title, performer) performer) 2. Partial display 2. Partial display (composer, title) (composer, title) 3. Minimal display 3. Minimal display (composer) (composer) 4. No display: number 4. No display: number of results of results Prototype II: New Query Dataset • 8556 mp3 files extracted from 714 albums by 315 different artists • Rock/pop and electronica • Labeled with English terms (by Eric Brochu) ALB Fever to Tell ART Yeah Yeah Yeahs REL Apr 29, 2003 GEN Rock • New query is an “extension” of old, linked by STY Indie Rock, Garage Punk line, colour, and position TON Cathartic, Exuberant, Boisterous, Passionate, Brittle PAT /cs/beta/SCRATCH/music/mp3library/Yeah Yeah • Old queries fade and shrink with time Yeahs/Fever to Tell Implementation Current status & Next steps • Architecture 7 8 9 10 11 12 13 Familiarize with database structure, refine prototype design – Flat (at the moment): since the amount of data 14 15 16 17 18 19 20 17 processing required is not extensive Implement basic layout and individual element selection 21 22 23 24 25 26 27 • Platform and language: Implement semantic zooming, F+C with animation – Java using Eclipse IDE on Windows 28 29 30 1 2 3 4 Implement new keyword query (spatial layout) • Libraries 5 6 7 8 9 10 11 – swt.jar Implement new keyword query (animation) – No other graphics library used (yet…) 12 13 14 15 Preparation of report and presentation 5
Demo 6
Recommend
More recommend
