Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Future Trendsin
Search UserInterfaces
Dr. Marti HearstUC Berkeley
i-Know Conference Keynote Sept 1, 2010
Forecasting the Future
First: What are the larger trends?
In technology?In society?
Next: Project out from these.
3
NaturalInterfaces
Wide adoption of touch-activated devices withexcellent UI design
Latent preferencesfor audio and video
Wide adoption ofmobile devices withdata service
Wide adoption ofsocial media anduser-generatedcontent
Improvements inNLP, speechrecognition viahuge data sets
Increasinglyavailable rich,integrated datasources
4
Trend: Natural InterfacesWhat does this mean for Search UIs? Longer, more natural queries
Including “sloppy commands”
Social Search Better Mobile Interfaces
Audio (spoken) queries and results
Context-sensitive results
Longer term: Video and audio dominating text
Dialogue / conversational interactions
Trend: More Natural Queries
6
Trend: Longer, more natural queries
The research suggests people prefer to state theirinformation need rather than use keywords. But after first using a search engine they quickly learned that
full questions resulted in failure.
Average query length continues to increase Major search engines are now handling long queries well.
Information worded as questions is increasing on theweb. From social question-answering sites and forums.
7
A recent example: keywords failed
8
A recent example: ask as a question
9
A recent example: get an answer
10
Advanced user interface design
11
Trend: More Natural Queries
Blend two ideas:
“sloppy commands”
predictions based on user behavior data
This is subtly and steadily increasing insophistication across many interfaces
12
Sloppy Commands
Like command languages, but
the user has a lot of flexibility in expression
so memorization is not required “time graz” “what time is it in graz” “graz time now”
13
Sloppy Commands + Visual Feedback
Can include rich visual feedback
Quicksilver in Apple
Inky by Miller et al.
14
Sloppy Commands + Rich Data
Combine Mozilla’s Ubiquity and Freebase tomake a flexible predictive query engine By spencerwaterbed: http://vimeo.com/13992710
15
Sloppy Commands + Rich Data
Combine Mozilla’s Ubiquity and Freebase tomake a flexible predictive query engine By spencerwaterbed: http://vimeo.com/13992710
16
Far Future Trend: Dialogue
We’re still far away.
SIRI is promising as a moveforward; based on state-of-the-art research.
Trend: Social Search
People are Social; Computers are Lonely.
Don’t Personalize Search, Socialize it!
18
Social Search
Implicit: Suggestions generated asa side-effect of search activity.
Asking: Communicatingdirectly with others.
Collaborative: Working withother people on a search task.
Explicit: knowledge accumulating viathe deliberate contributions of many.
19
Trend: Social Search
Social ranking Explicitly recommended
Digg, StumbleUpon
Delicious, Furl
Google’s SearchWiki
Implicitly recommended Click-through
People who bought…
Yahoo’s MyWeb (now Google Social Search)
20
Social Search: Implicit Suggestions
Human-generated suggestions still beat purelymachine-generated ones
Spelling suggestions
Query term suggestions (“search as you type”)
Recommendations (books, movies, etc)
Ranking (using clickthrough statistics)
21
Social Search: Asking for Answers What do people ask of their social networks?
Type % Example
Recommendation 29%Building a new playlist – any ideas for goodrunning songs?
Opinion 22%I am wondering if I should buy the Kitchen-Aidice cream maker?
Factual 17%Anyone know a way to put Excel charts intoLaTeX?
Rhetorical 14% Why are men so stupid?
Invitation 9% Who wants to go to Navya Lounge this evening?
Favor 4% Need a babysitter in a big way tonight…anyone??
Social connection 3%I am hiring in my team. Do you know anyonewho would be interested?
Offer 1% Could any of my friends use boys size 4 jeans?
Morris et al., CHI 2010
22
Social Search: Explicit Help viaQuestion-Answering Sites
Content is produced in a manner amenableto searching for answers to questions.
Search tends to work well on these sites andon the internet leading to these sites
Like an FAQ but with many authors, and
with the questions that the audience really wants theanswers to, and
written in the language the audience wants to use.
23
Social Search: Seeing what peopleyou know have seen Yahoo MyWeb, Google Social Search
24
Social Search: Explicit SuggestionsBuilding Knowledge Social knowledge management tools seem promising
Utilize the best of social networks, tagging, blogging,web page creation, wikis, and search.
Millen et al., CHI 2006
25
Social Search: People Collaborating
Tools to help with this are only just beginning.
Pickens et al., SIGIR 2008
26
Future Trend: The Decline of Text
Or: the rise of audio/video The “cultural heavy lifting” in America is moving from
text to audio and video. Video and audio are now easy to produce and share.
Pew: Use of video sharing sites doubled from 2006-2009
YouTube: Video “responses” apparently arose spontaneously
Videos for presidential debates were mundane.
Millions of video views; no where near this number for articlereadings
Pew: Marketing emails with podcasts 20% more likely to beopened.
Movies with subtitles do poorly in the U.S.
Newspaper web sites are starting to look like TV.
(Full essay at http://edge.org/q2009/q09_9.html#hearst)
27
Future Trend: The Decline of Text and Riseof Audio/Video
Other advantages Doesn’t require literacy (but can serve as a bridge)
Robust video devices to teach agriculturaltechniques
What about the popularity of texting? An interesting counter-trend
Especially popular among youth Cheap
Can be done surreptitiously
(Full essay at http://edge.org/q2009/q09_9.html#hearst)
28
Future Trend: The Decline of Text and Riseof Audio/Video
What is currently solved: Ease of creating and sharing videos
What needs improvement Editing tools
The main impediments to more fully replacing text are: The need for better search and scanning of audio and video
A means for silent audio input
What about the popularity of texting? An interesting counter-trend
Especially popular among youth Cheap
Can be done surreptitiously(Full essay at http://edge.org/q2009/q09_9.html#hearst)
29
Future Trends … not so much?
Personalization
Visualization … some breakthroughs areneeded.
30
Summary
As CS gets more sophisticated, we can buildsearch interfaces that allow people tointeract more naturally: More language-like queries
Speaking & hearing rather than typing & reading
More able to interact with other people whiledoing search tasks
More able to use the knowledge in peoples’ heads
Thank you!
Marti HearstUC Berkeley
Book full text freely available at:
http://searchuserinterfaces.com