A Comparison of Consecutive and Concurrent Input Text entry.ppt

《A Comparison of Consecutive and Concurrent Input Text entry.ppt》由会员分享，可在线阅读，更多相关《A Comparison of Consecutive and Concurrent Input Text entry.ppt（30页珍藏版）》请在麦多课文档分享上搜索。

1、A Comparison of Consecutive and Concurrent Input Text entry Techniques for Mobile Phones,Daniel Wigdor & Ravin Balakrishnan,2,Text Messaging,Estimated 500,000,000,000 text messages in 2003 worldwideMore popular outside North America,3,Ambiguity,Pressing “2” : 2,a,b,c,A,B,C,4,Problem,Multiple selecti

2、on actions required MacKenzie & Soukoreff: Between group selection: 2,a,b,c,A,B,C Within group selection: a,5,Consecutive Selection,Most prior techniques consecutive: First make between group selection Then make within group selection, or disambiguation of automatic selection MultiTap, T9, 2-Key, Le

3、tterWise, WordWise What about selecting concurrently?,6,Concurrent Selection: TiltText,Presented at UIST 2003 Between-group selection: press key Within-group selection: tilt phone Actions performed concurrently Suggests new area of research,7,Taxonomy: Multiplexed Keypad,1,2 not possible 4 & 5 are r

4、egressive 3 suggests a use for chording,8,Mobile Chording,Chording input for mobile devices:Selection tasks suggests better fit for chording on mobile phone,9,ChordTap,Mobile phone selection tasks: Between group selection (phone keypad) Within group selection (chord keys) Performed concurrently,10,D

5、esign Issues,Mapping chord states to within-group selection Event that triggers text generation,11,Chord Mappings,Ignoring case, buttons have 4 or 5 characters Each chord has 2 states (down=0, up=1) log25 = 3, 3 chords needed,12,Chord Mappings,13,Character Generation Event,Issue: when is character g

6、enerated? Actions To Generate a Character:,Depress keypad button,Depress chord Key,1.,2.,3.,Release which?,14,Why Does it Matter?,Non-event keys can be visualized Reduces erroneous text entry Helps with learning Non-event key can be held for subsequent character, savings,15,Keypad Button Release (eg

7、):,1.,Depress keypad button:,2.,Depress chord key:,Screen:,3.,Release keypad button text is generated. If next character requires same chord, keep it held down to skip step 2.,16,Work Savings,Subsequent characters with same chords:,17,Keypad Button Release,Allows for visualization of within-group se

8、lection Allows same chord to be held for successive characters Savings on 20% of pairings,18,Chord Key Release (eg):,1.,Depress chord key:,2.,Depress keypad button:,3.,Release chord key text is generated. If next character requires same chord, keep it held down to skip step 2.,19,Work Savings,Subseq

9、uent characters with same keypad button:,20,Chord-Key Release,Character generated when chord-key released Allows for visualization of between-group selection Allows same button to be held for successive characters Savings on 9% of pairings,21,Chord-Key or Keypad Button,Text generated every time any

10、key released Allows for no pre-visualization Greatest savings: 29%,22,Our Prototype,Button release text generation Equipped with 3 chords Implemented on Mot i95cl Mouse board for chords 2-handed,23,The Study,Comparing ChordTap to MultiTap Between-Subject Design: 15 participants 3 techniques (MultiTa

11、p: 1 or 2 handed, ChordTap) 16 blocks of 20 phrases each 2 sessions Same phrases for both techniques Measured time & accuracy Participants told to correct mistakes,24,Results: Overall Speed,End of experiment: MT1: 11.05, MT2: 12.04, CT: 16.06,Day 1,Day 2,25,Power-Law Extrapolation,26,Results: Error

12、Rate,Day 1,Day 2,27,Conclusions,Identified taxonomy of research Explored a new area within taxonomy Fit chording to mobile-phone keypad: Mappings Text generation event Implemented prototype Formal study conducted Chording better than MultiTap,28,Future Work,Chording for one hand Further exploration of taxonomy Study of text generation event and mappings,29,Acknowledgements,Tovi Grossman Maya Przybylski Krista Strickland DGP Lab members Study participants Microsoft Research,30,Thank You!,