CSE 450

1
CSE 450 Web Mining SeminarProfessor Brian D.
DavisonFall 2005

• A Presentation on
• When Experts Agree Using Non-Affiliated Experts
  to Rank Popular Topics
• K. Bharat G. A. Mihaila
• WWW10 Conference, May 2001, Hong Kong
• by
• Osama Ahmed Khan
• 10/06/2005

2
Problem

• Query on Popular Topic
• Content Analysis

Solution

• Most Authoritative Pages

3
Technical Terms

• Expert
• Recommendation
• Non-affiliation

4
Hilltop Algorithm

• Expert Lookup
• Detecting Host Affiliation
• Expert Selection
• Expert Indexing
• Target Ranking
• Computing Expert Score
• Computing Target Score

5
Detecting Host Affiliation

• Conditions
• Same first 3 octets of IP
• 127.0.0.1 127.0.0.15
• Same rightmost non-generic token of hostname
• www.ibm.com www.ibm.co.mx
• Union-Find Algorithm

6
Expert Selection

• Retrieve all webpages with
• Out-degree gt Threshold (k)
• (e.g. k 5)
• Expert will have
• URLs pointing to k distinct non-affiliated hosts

7
Expert Indexing

• Inverted Index
• Mapping Keywords to Experts
• Key Phrases
• Match Positions

8
Computing Expert Score

• Condition
• Atleast 1 URL with all query keywords
• Expert Score (S0, S1, S2)
• Si SUMkey phrases p with k-i query terms
  LevelScore(p) FullnessFactor(p,q)
• Expert_Score 232 S0 216 S1 S2

9
Computing Target Score

• Condition
• Atleast 2 non-affiliated experts
• Target Score
• Edge_Score(E,T) Expert_Score(E)
• SUMquery keywords w occ(k,T)
• Target_Score SumEdge_Score(E,T)

10
Evaluation

1. Locating Specific Popular Targets

11
Evaluation (Contd.)

1. Gathering Relevant Pages

12
Conclusion

• Characteristics
• Popular Queries
• Expert Subset
• Hilltop vs.
• PageRank
• Topic Distillation

13

• Thank You