Technology

1
Technology

Intermediation
Jerry Waller and Lee Smith

• INLS 180
• Human Information Interaction
• The University of North Carolina at Chapel Hill,
  Chapel Hill, NC

2
Introduction

• Interface designers are becoming more aware of a
  crucial missing factor within the design of many
  information systems a personal or human touch.
  (Brown Duguid 2000)

3
What is Intermediation?

• The act or process of serving as a intermediary.
• In the case of technological intermediation,
  intermediaries are essentially proxies,
  workingon behalf of others. (Brown Duguid
  2000)
• Intermediaries are programs that operate on
  information flowing between its producer and its
  consumer. (Maglio and Barrett 2000)

4
Intermediaries Personalize Information Streams

• (Paul Maglio Rob Barrett 2000)

5
Classifications of Intermediaries

• Where they run and who controls them
• Quick dissemination, discovery and retrieval
• By function
• Customization
• Considers users information in relation to
  environment
• Filtering
• Annotating

6
Classifications cont.

• Transcoding
• e.g. HTML to WML
• Aggregating
• Merging results into a single document
• Caching
• History/Monitoring -- data last stored or altered
• Adds element of organization to search

7
Examples

• Marketing agents
• (WBI) Web Intermediaries
• On the Web at Home

8
Other Intermediary Examples

• Autonomous agents bots
• Track personal preferences
• Makes history-based suggestions to user
• Redundancy

9
WBI

• Composed of MEGs
• Monitors
• Editors
• Generators
• Adds intermediary functions to the Web
• WBI is itself a proxy server
• Can transform information at server, client

10
Opportunities for Improvement

• Moores Law?

11
Agents and AngelsOR, Religious Allegory Goes
Hi-Tech

• Agents deal withcascades of information and
  promise to handle many intricate human tasks.
  (Brown and Duguid 2000)
• Monitoring actions/search queries
• Correction of human errors

12
Agents

• Chatterbots simulate text-based human responses.
• Agents help manage the Web.
• Google (wasnt included in article)
• Yahoo
• Excite
• Alta Vista
• Agents They know who you are!
• You dont know that they are dogs.

13
Agents on Our Shoulders

• Anthropomorphized
• Human names
• Human job titles
• make humans seem like bots
• (Brown and Duguid 2000)
• Human and Technology gap narrowing

14
Fallen Agents

• Agents lack the capability to implement judgment
  and discretion
• Intel/Symantec/Zero-Knowledge Program
• Hostile code (Contained viruses)
• Revealing of flaw in Intels chip

15
Information Brokering

• Brown Duguid postulate that information
  brokering agents may make classical information
  providers redundant.
• Far from perfect
• Sherlock
• Data mining

16
Product Brokering

• Powerful and useful, but
• The agents are subject to manipulation
• Amazon.com
• SAABRE

17
Merchant Brokeringand Its Concerns

• long-distance marketswork best with
  standardized measures and standardized products.
  (Brown and Duguid 2000)
• May lead to less choice
• Service and quality suffer, because bots
  dontcantcare.
• No reasoning capabilities or consciousness

18
On the Web at Home

• (ELIS) Everyday Life Information Seeking
• Great definition of context
• Context can be construed as those information
  environments within which information behaviors
  take place. (Rieh, Soo Young 2004)
• Concept of the hanging out space.
• Search engines were not the first place the
  subjects looked for information.

19
Table 1.
(Rieh, Soo Young 2004)
20
Table 3.
(Rieh, Soo Young 2004)
21
Table 4.
(Rieh, Soo Young 2004)
22
Most Important Issues

• Potential biases
• Environment where study was conducted
• Requirement of Broadband
• Sponsor of the study Excite
• Think of the SAABRE system
• How reliable/viable are results considering the
  restraints on the study?

23
Study Results

• Always on web aspect
• Feeling Successful attitude/mentality
• (Rieh, Soo Young 2004)
• Bot Development needs more research
• Some researchers feel that until the elementof
  morality can be included in initial bot
  infrastructure they will continue to lack what
  makes us human.

24
Wonder of Wonders!

• The Internet substantially influenced information
  seeking at home.
• The Web has become embedded in everyday life.

25
As it Now Stands

• Intermediaries are only able to follow
  instructions literally.
• A literal-interpretation model is not capable of
  making adaptations when necessary.
• Judgment and discretion are not features of
  software. (Brown and Duguid 2000)

26
More Disadvantages

• Lack capability to implement judgment and
  discretion into situations
• Negotiation question of supply and demand
• (Brown and Duguid 2000)
• Bot Foibles
• Possible manipulation/tampering with
  dataMalicious activity/misrepresent-disruptive

27
Common Threads

• All dealt with personalization or better
  customization of agents/user preferences.
• The invisible hand of the agents

28
Future Strategies

• Create agents that anticipate users needs
• Bill Gates confidently predicts that bots will
  soon develop personalities.
• (Brown and Duguid 60)
• Better encryption and less restriction (Moores
  law)

29
Conclusions

• A viable system must embrace not just the
  technological system, but the social systemthe
  people, organizations, and institutions
  involved. (Brown and Duguid 60)
• More research should be done regarding home-based
  information seeking and web searching.
• The notion of intermediaries can be generalized
  to incorporate processes that produce
  personalized information and much more. (Maglio
  and Barrett 2000)

30
God is in the Details

• Better intermediaries will require a better
  understanding of
• The contribution of the social fabric
• The role of human restraint in the functioning
• of the invisible hand
• The computer-literacy (or lack thereof ) of the
  user

31
Definitions

• Chatterbots or bots
• Bots
• Portal
• Wetware
• Agents
• Invisible Hand
• Intermediary
• ELIS
• Feeling successful

• Software programs that simulatehuman responses

Proxies that act on behalf of others and not
themselves
Points of origin where people plan and begin
voyage on Net
Individual information processing by
goal-pursuing agents
Electronic trackers/compliers of user preferences
Adam Smiths description of unseen market forces
Program/agent that meaningfully transform
information
Term used by Soo for every day life information
seeking
Emotion users receive when search is effective
32
Group Questions

• How can a fair play factor be ensured when
  using intermediaries?
• Is Moores Law in the context of intermediariesa
  fallacy?
• Compare Human foible factor to Bot foible
  factorWhich is better? Worse? Why?