Networked Information Resources - PowerPoint PPT Presentation

About This Presentation
Title:

Networked Information Resources

Description:

Social matching (match.com, expertise recommender) Collaborative filtering ... E.g. Amazon, Netflix. Steps in collaborative filtering. User enters profile of ratings ... – PowerPoint PPT presentation

Number of Views:12
Avg rating:3.0/5.0
Slides: 10
Provided by: Supe1
Category:

less

Transcript and Presenter's Notes

Title: Networked Information Resources


1
Networked Information Resources
  • Word-of-mouth collaborative filtering

2
Impacts of the Internet
  • Ownership vs. access
  • Concurrent, ubiquitous access
  • End-user searching
  • Information literacy
  • Organization of networked resources
  • Proactive role in knowledge production process
  • New knowledge discovery tools
  • Blurring of institutional boundary

3
Social production of knowledge
  • The way a publication contributes to knowledge,
    if it does, is by changing the collective opinion
    of the specialist group. The size of a
    contribution is measured by the size of the
    change produced in the collective opinion.
    Producing knowledge is not something the
    individual can do. All one can do is make
    proposals to the group and hope that they will be
    accepted in some form. If they are, and if the
    proposal is accepted by the group as settling
    some question for the time being, then a crucial
    step toward a contribution of knowledge has been
    made. (Wilson 1983, p.48)

4
Information Filtering (IF)
  • SDI (Selective Dissemination of Information)
  • Google Alerts

5
Differences between IF IR
  • 1. long-term vs. short-term needs
  • 2. Incoming stream of data vs. static collection
  • 3. Removal vs. finding

6
Automatic recommender system
  • Content-based
  • Collaborative filtering
  • Social matching (match.com, expertise recommender)

7
Collaborative filtering
  • Also referred as social recommendation system
  • E.g. Amazon, Netflix

8
Steps in collaborative filtering
  • User enters profile of ratings
  • System locates people with similar profiles
    (neighbors)
  • Neighbors ratings are combined to form
    recommendations

9
Social cognitive issues
  • Trust
  • Risk
  • Cold start
  • Privacy
Write a Comment
User Comments (0)
About PowerShow.com