blended libraries (harald reiterer)
Post on 27-Jan-2015
119 Views
Preview:
DESCRIPTION
TRANSCRIPT
h"p://hci.uni-‐konstanz.de
Prof. Dr. Harald Reiterer University of Konstanz
Chair for Human-Computer Interaction Harald.Reiterer@uni-konstanz.de
hci.uni-konstanz.de
University of Konstanz
Blended Libraries
Future user interfaces concepts to support informa0on seeking and
collabora0on
2
Introduc7on • University of Konstanz, Department
of Computer & Informa0on Science, Head of the Research Group Human-‐Computer Interac0on (since 1997)
• Research Topics: Interac0on Design, Usability Engineering, Informa0on Visualiza0on
• Main Goal: Improving the naturalness of the interac7on between human and computer
– Mul0modal Interac0on – Reality-‐based Interac0on
3
UIs for Informa7on Seeking Systems (ISS)
• INSYDER (ISS for the Web) – EU Project Eureka No. 29232 (1999-‐2001)
• INVISIP (ISS for Geo-‐Data) – EU Project No. IST-‐2000-‐29640 (2002-‐2004)
• Mobile Informa7on Visualiza7on (ISS for mobile devices) – DFG PhD Graduate Program (2005-‐2007)
• MedioVis (ISS for digital/mul0media libraries) – DFG LIS 4-‐54281(3) (2007-‐2009)
• Blended Library -‐ Ministerium für Wissenschaa, Forschung und Kunst, Baden-‐Würcemberg (2011-‐2013)
more informa0on: hcp://hci.uni-‐konstanz.de
4
Mo7va7on Poor Usability of OPACs (2007)
5
MedioVis Demo of System in Use
Integration of Multimedia Metadata and Web Services
6
Library of the Future – Blend Physical & Digital Library
digital library physical library
status-‐quo co-‐existance
vision co-‐opera0on
7
Blended Library – Domains of Design Individual Interac7on: Blending real world interac0on & objects (e.g. scribbling text, sketching objects, grasping tokens) with digital ac0ons & representa0ons (e.g. to move digital books or movies, touch items of a facet token value wheel). Social Interac7on & Communica7on: Different user can interact at the same 0me on an equal manner considering real-‐word social conven0ons of communica0on (e.g. mul0-‐touch displays allows mul0user interac0on; tokens could be grasped equally). Workflow: Fluid change between real-‐world workflows and computer-‐supported workflows (e.g. search in a bookshelf – search on a Tabletop). Physical Environment: Blending the power of real-‐world devices, furniture, rooms & buildings with the power of virtual means (e.g. combining tables, chairs, walls, floor, ceiling, light, sound with mul0-‐touch planar, spherical, cylindrical, curved displays); new arrangements of place and space.
8
What kind of UI for different Displays? Large – small
Different Input modali0es (e.g. Touch, Stylus, Keyboard, Mouse)
Real-‐0me Synchroniza0on
?
9
Zoomable Object-‐Oriented Informa7on Landscape
10
Query by Real World Objects, e.g. DVD
11
Query by Real World Objects, e.g. DVD
12
Blended Library – Coopera7ve Search with Search Tokens
13
Blended Library – Coopera7ve Search with Search Tokens
14
User Study -‐ Se_ng
Par7cipants: 75 par0cipants, tested in 25 randomly selected 3-‐person-‐groups (triads) Between-‐Subject Design (12 + 13 groups) Two UI Condi7ons: -‐ Synchronized PC Interface -‐ Search token UI: Physical search token + Mul0-‐touch Tabletop
Research Ques7on: What impact has a search token UI on the • Interac0on Strategy, • Communica0on, • Roles of Collabora0on?
15
Selected Results
• Interac7on: Tokens featured more parallel interac0on. • Interac7on: Tokens enhance the coopera0on between the par0cipants, they
even share their tokens! • Verbal Communica7on: no significant differences. • Non-‐Verbal Communica7on: With the Search Token UI gestures and posture
were perceived without needing to look up from the display. • Roles of Collabora7on: 5 different roles iden0fied. Search Token UI allowed
ac0ve lead users to take on a more dominant role within the groups. Even the passive group members showed a more ac0ve role compared with the passive users in the synchronized PC condi0on.
16
Tabletops for Co-‐Located Collabora7ve Faceted Search
17
Tabletops & Collabora7ve Search: Why? • poten0ally great support of mechanics of collabora7on
[Gutwin & Greenberg 2000]
• closer face-‐to-‐face collabora7on and more equitable working style [Rogers & Lindlay 2004]
• increased awareness and be"er group work experience [Amershi & Morris 2008]
18
Facet-‐Streams Approach • Hybrid Surfaces [Kirk et al. 2009]
– Combining mul0-‐touch interac0on with tangible interface elements
• Faceted Search [Hearst 2009] – itera0vely filtering the whole informa0on space based on metadata, instead of popula0ng a result set based on keywords
19
Facet-‐Streams Approach
Visual & tangible filter/flow metaphor for Boolean logic Redesign of [Young & Shneiderman 1993, Hansaki et al. 2006] for tabletops
20
Research Ques7ons • Does Facet-‐Streams turn collabora0ve product search into a
fun and social experience with increased group awareness?
• Can Facet-‐Streams support the great variety of different search strategies and collabora0on styles in different teams with a simple but flexible design?
• Can Facet-‐Streams harness the expressive power of facets and Boolean logic without exposing users to complex formal nota0ons?
21
Study 1 -‐ Facet Streams in collabora7ve use • 72 par0cipants in 24 groups of three • Students and faculty staff from a variety of fields (only 2
computer science students)
22
The Web-‐Interface
23
Task • Goal: Facilitate compromise nego,a,on in a controlled
environment
24
Results -‐ Strategies
25
Results – Study 1 • Facet-Streams equally effective as established Web
designs for faceted navigation (although it introduces novel and unfamiliar hybrid interaction techniques and visual metaphors).
• Users perceived using Facet-Streams as a fun experience and considered its design as innovative.
• Increased awareness and better mutual support among collaborators was observed.
• Variety of different search strategies and collaboration styles could be realized with our design.
• Seamless transitions between tightly-coupled collaboration and loosely-coupled parallel work.
26
Study 2 – Comprehensiveness of Boolean logic
• 7 par0cipants (no computer science students) • Individual sessions (no group interac0on) • Focus:
– Task 1: Can people understand and learn our visual representa0on of Boolean logic (without being made aware that this is Boolean logic!)?
– Task 2: Are they able to formulate a complex and evolving query based on natural language instruc7ons?
27
Study 2 -‐ Task 1 3 networks 3 Hotels per network
4 spots per hotel
=36 Ques0ons
Can people understand and learn our visual representa0on of Boolean logic (without being made aware that this is Boolean logic!)?
28
Results Study 2 – Task 1
Correct Wrong
Total 234 (92,9%) 18 (7,1%)
AND 101 (96,2%) 4 (3,8%)
OR 133 (90,5%) 14 (9,5%)
29
Study 2 – Task 2 • Are people able to formulate a complex and evolving query based on
natural language instruc7ons?
• Experimenter: Role of a costumer – User: Role of a travel agent
– „I would be interested in a hotel with a price between 50 and 120 EUR, a very high room quality, and at least 4 stars.“
– „It should be in Spain.“ – „Lets include those hotels, that have a high loca0on quality instead of
a high room quality as an alterna0ve. They should s0ll have the other features.“
– … – Requirements: Zero Hit, AND, OR, temporary deac0va0on, mul0ple
output streams, parallel streams
30
Results Study 2 -‐ Task 2 • 6 of 7 correctly integrated the OR connec0on
• Boolean: (room quality = 4-‐6 OR loca,on quality = 5-‐6) AND (hotel stars = 4-‐5) AND (country = Germany OR country = Spain) AND (features = Bar+Pool)
31
Results – Study 2 • Users were able to quickly learn and apply our visual metaphor for Boolean logic
• Users succeeded in formula0ng complex Boolean queries based on natural language instruc0ons
32
Natural User Interfaces “The goal is to make computer interaction more like interacting with the real, non-digital world.” Reality-‐Based Interac0on (Jacob et al. 2008)
User flicks screen to left or right to see next Album – all objects in the real word have spatial relationships.
Fast flick will keep contacts scrolling after finger has been removed, as if list has mass.
Natural gestures control the applications.
User do not share one single input device – Tabletop allows to draw more directly on existing social interaction skills.
33
Conclusion • Conclusions • Blended Library draw strength by building on users’ pre-existing
knowledge of the everyday, non-digital library world. • It employ experiences of reality and makes the interaction with the
computer much more natural. • It blends the benefits of digital libraries with the benefits of physical
libraries to make visiting and using a library a new experience.
[Microsoft Research UK 2008, Being Human: Human-Computer Interaction in the Year 2020]
34
Papers and Videos available on our Website hci.uni-‐konstanz.de
Ques7ons?
top related