Phlat - PowerPoint PPT Presentation

1 / 24
About This Presentation
Title:

Phlat

Description:

24559.00 15852.00 4529.00 8114.00 Date. Rank. Default UI. Number of Queries. Sorting Usage. Title: Phlat: Author: edcutrell Last modified by: Edward Cutrell Created Date: – PowerPoint PPT presentation

Number of Views:17
Avg rating:3.0/5.0
Slides: 25
Provided by: edcut
Category:
Tags: phlat

less

Transcript and Presenter's Notes

Title: Phlat


1
Phlat
  • Phast phlexible personal search and organization

Ed CutrellMicrosoft ResearchAdaptive Systems
Interaction Group
2
Outline
  • Personal search today
  • Search with Phlat
  • With Susan Dumais, Daniel Robbins, Raman Sarin
  • Search/Browse/Filter
  • Tagging
  • Evaluation Lessons
  • Memory Landmarks
  • Implicit Query

3
Search in 2004
  • Many locations, interfaces for finding things
    (e.g., web, mail, local files, help, history,
    notes)

the No.1 question we're trying to solve in
Longhorn is Where's my stuff? Right now,
file space on any PC is a cesspool. Bill
Gates, FORTUNE interview, June 23, 2002
  • Often slow

4
Stuff Ive Seen
  • In 2001, we built SIS
  • Deployed to 3000 people in MSFT
  • Allowed us to study how people use ( want to
    use) personal search
  • Limited prototype but very successful

5
Search Today
  • Desktop search hits the bigtime!
  • MSN ToolbarWindows Desktop Search
  • Google DS
  • Yahoo! Toolbar (X1 DS)
  • Copernic DS
  • Apple OS X Tiger with Spotlight
  • Many others
  • Unified index of full-text metadata for
    different stores
  • Re-use vs. initial discovery

6
Simple UI on Rich Client
7
Search with Phlat
  • Phlat asks Can we design an intuitive and
    flexible interface that replaces the traditional
    search/browse dichotomy?
  • Topple the tyranny of the rigid hierarchical file
    system!(folders keep the metadata and toss the
    metaphor)
  • Embrace the power of the index!(real-time
    interaction on any axis)
  • Powered by Windows Desktop Search
  • Phlat is a client shell that runs against the
    index built maintained by Windows DS (MSN
    Toolbar)
  • Queries routed through SIS Communications layer
    (C)

8
Faceted property filtering
  • 5 canonical properties to filter on (extensible)
  • Prop filters integrated with query
  • Type-in for each property (wordwheel)
  • SearchBrowse

9
Tagging
  • Apply single set of user-generated metadata to
    all files, email, etc.
  • Allow but not require hierarchy
  • Tags are directly associated with files (NTFS or
    MAPI props)
  • Need to hook CFD
  • Delay associated with query indexer

10
Demo
11
Phlat Demo Points
  • Full-text property search rendered similarlyno
    stuck filters, browse/search continuum
  • Sort and filter resultsstable sort, static
    headers
  • Wordwheel for properties--hierarchy
  • Tagging (hierarchy auto-open to level for
    tagged items)
  • ?Right-click and other advanced functionality
  • ?Saved queries, Standing queries
  • ?MyTopicsArun Surendran

12
Evaluating Phlat
  • Internal deployment
  • 500 downloads
  • Users include program management, test, sales,
    development, administrative, executives, etc.
  • Research techniques
  • Free-form feedback
  • Questionnaires Structured interviews
  • Usage patterns from log data
  • Gaze tracking studies
  • Lab studies for richer UI (e.g., timeline, trends)

13
User Logging
  • Phlat records a broad range of query
    characteristics to help us understand the
    characteristics of usage.
  • Age of items opened
  • Characteristics of query and filters used
  • Number of tags applied
  • Cycle of query iteration
  • Special instrumented version for gaze tracking
    research

14
Phlat Observations 1(unified access)
  • Metadata quality is variable
  • Email rich, pretty clean
  • Web little, not very useful for retrieval
  • Files some, but often wrong
  • Human annotation dont depend on it( but good
    UI in authoring environments can help)
  • Memory depends on abstractions
  • Date is dependent on the object!
  • Appointment, when it happens
  • File, when it is changed
  • Email and Web, when it is seen
  • People
  • To, From, Cc, Author, Artist
  • Document names

15
Phlat Observations 2(tricky business)
  • Internet cache web history are very useful but
    tricky areas
  • Cache can be confusing, but re-access is
    important
  • Access to temp files is GREAT
  • Tagging
  • Hierarchy important for tag organization and
    cognitive assistance, but flexibility is key
    (think organic)
  • MUST make tagging UI ubiquitous and available at
    inflection points of information consumption
  • Sharing tricky

16
Timeline w/ Landmarks
  • Timeline interface
  • (with Merrie Ringel Eric Horvitz)
  • Augmented with landmarks as pointers into human
    memory
  • General holidays, world events
  • Personal important photos, appointments
  • Heuristics or Bayesian models to identify
    memorable events

17
SIS, Timeline w/ Landmarks
18
Demo
19
Timeline Experiment
With Landmarks
Without Landmarks
20
Contextualizing Search
  • Search is not the end goal
  • Finding
  • Always available search from task bar
  • Search from within apps
  • Plug-ins w/ queries to Phlat/Desktop Search
  • Research pane
  • Implicit queries
  • Proactively finding results
  • Using
  • e.g., drag/drop, tag, etc.

21
Implicit Queries
Quick searches for people associated with the
message and Subject.
Background search on top k interesting terms from
message, based on users index Score tfdoc /
log(tfcorpus1)
Go to SIS for immediate detailed search. Box
autofills with IQ search terms.
Top N hits for IQ based on size of window. Open
items directly.
22
IQ Demo
23
Learn More
  • Phlat is usable NOW! http//phlat
  • Design goes well beyond the prototype. Feel free
    to view spec http//team/sites/phlat
  • More info, papers, etc.
    http//research.microsoft.com/cutrell
  • Related projects at MSFT
  • Memex, Tesla, Longhorn

24
Thank You!
25
(No Transcript)
26
SIS Usage Data
  • Detailed analysis for 234 people, 6 weeks usage
  • Personal store characteristics
  • 5k 100k items
  • Query characteristics
  • Short queries (1.6 words)
  • Few advanced operators or fielded search in query
    box (7)
  • Frequent use of query iteration (48)
  • 50 refined queries filtered type, date (most
    common)
  • 35 refined queries changed the query
  • 13 refined queries re-sorted
  • Query content
  • Vs. Spink et al.s analysis of web queries
  • Importance of people
  • 29 of the queries involve peoples names

27
SIS Usage Data, contd
  • Characteristics of items opened
  • File types opened
  • 76 Email
  • 14 Web pages
  • 10 Files
  • Age of items opened
  • 7 today
  • 22 within the last week
  • 46 within the last month
  • Ease of finding information
  • Easier after SIS for web, email, files
  • Non-SIS search decreases for web, email, files
Write a Comment
User Comments (0)
About PowerShow.com