Surviving the Information Explosion

1
Surviving the Information Explosion

• Christine Alvarado and Jaime Teevan

2
Overview

• Motivation
• Background
• Our study
• Preliminary results
• Future work

3
Let Us Interview You!

• Email

• Whats the last email you read? What did you do
  with it?
• Have you gone back to an email youve read before?

• Files

• Whats the last file you looked at? How did you
  get to it?
• Have you searched for a file?

• Web

• Whats the last Web page you visited? How did you
  get there?
• Have you searched for anything on the Web?

4
The Information Explosion

• You must extract information from
• 1.6 billion web pages Google
• Dozens of incoming emails daily
• Hundreds of files on your personal computer

5
Limited Organizational Tools
6
Limited Organizational Tools

• Many separate tools
• Limited organizational support
• Organizational burden on user
• Information overwhelms tools

7
HaystackPersonal Information Storage
Web pages
Email
Files
Calendar
Contacts
8
HaystackPersonal Information Storage
What was that paper I read last week about
Information Retrieval?
Haystack
9
HaystackPersonal Information Storage
Ah yes! Thank you.
Haystack
10
User Interface
Microsoft Outlook
Pine
11
User Study Goals

• Search
• Frequency
• Type
• Organization
• Patterns
• Use

12
Pre-Study Summer 2001Setup

• 6 subjects
• Observed/recorded working for 1-2 hours
• Follow-up interview

13
Pre-studyAreas to Explore

• Window placement
• Desktop organization
• Context switches
• Navigation
• Searches

14
Previous Work

• Paper documents
• Malone, 1983, Whittaker Hirshberg, 2001
• Files
• Barreau Nardi, 1995
• Web (bookmarks)
• Abrams, 1998
• Email/Calendar
• Whittaker Snider, 1996, Bellotti Smith,
  2000

15
Whittaker and Hirshberg, 2001

• Method
• Web survey, 50 ATT employees
• Follow-up interview, 14 employees
• Goal
• Determine attitudes toward paper information
  organization
• Results
• Obsolescence
• Uniqueness
• Filers vs. Pilers

16
Method

• Subjects
• 15 MIT CS graduate students (5 women, 10 men)
• Setup
• 10 short interviews ( 5 min.)
• 1 long interview ( 45 min.)
• Topics
• Web, Email, Files

17
Short Interviews

• 2 question types
• What was the last email/file/web page you looked
  at?
• Did you search for any email/file/web page?
• Goal Discover patterns in searching and browsing

18
Long Interviews

• Guided tour of subjects bookmarks, email, and
  file system
• Goals
• Discover organizational patterns
• Relate organization to
• search/browse behavior
• Discover problems in
• organizational structure

19
Remember Your Answers?

• Results based on 85 short interviews

• Getting to a Web page

• Using a bookmark 57 of accesses

• Typing a URL 20 of accesses

• 19 of above followed links from there

• 3 out of 13 Web searches are for information
  that the user has seen before

• 64 of searched for email is found in the users
  Inbox

20
Results

• Quantitative
• Numbers, counts
• Reproducible
• Qualitative
• Anecdotes
• Building hypotheses
• Categorization of behaviors

21
Search Preliminary Results

• Different types of searches
• Directory lookup
• Confirming information exists
• Finding a specific piece of information (QA)
• Learning about a topic (Browse)
• Cross type searches
• Interactions with people
• Searching heavily relied on, very successful

22
Search Future Work

• Causes of failure
• Previously viewed information
• Additional cues used for retrieval
• Function of browsing during search

23
Organization Future Work

• Consistency of organization across types
• Context used in organization
• Organizations effect on search

24
Haystack Applying What We Learn

• Verify our conclusions
• Boundaries between information types
• Automation versus support
• Interaction between search and browsing

25
Questions?
Contact us with comments - calvarad_at_ai.mit.edu
- teevan_at_ai.mit.edu
To learn more about Haystack http//haystack.lcs.
mit.edu