SlideShare una empresa de Scribd logo
1 de 42
Descargar para leer sin conexión
Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                     The Choice of a Damping Function for
Notation
                         Propagating Page Importance
Rewriting
PageRank                    in Link-Based Ranking
Functional
Rankings

Algorithms

Comparison
                  Ricardo Baeza-Yates1 , Paolo Boldi2 and Carlos Castillo3
Conclusions
                               1. Yahoo Research – Barcelona, Spain
                                   2. Universit` di Milano – Italy
                                               a
                            3. Universit` di Roma “La Sapienza” – Italy
                                        a


                                     February 6th, 2005
Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
                  1   Notation
   C. Castillo

Notation          2   Rewriting PageRank
Rewriting
PageRank

Functional
Rankings
                  3   Functional Rankings
Algorithms

Comparison        4   Algorithms
Conclusions


                  5   Comparison

                  6   Conclusions
•›››››››››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting
PageRank          Let PN×N be the normalized link matrix of a graph
Functional
Rankings              Row-normalized
Algorithms            No “sinks”
Comparison

Conclusions
••››››››››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation          Definition (PageRank)
Rewriting
PageRank
                  Stationary state of:
Functional
Rankings                                        (1 − α)
                                         αP +           1N×N
Algorithms                                         N
Comparison

Conclusions
••››››››››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation          Definition (PageRank)
Rewriting
PageRank
                  Stationary state of:
Functional
Rankings                                        (1 − α)
                                         αP +           1N×N
Algorithms                                         N
Comparison

Conclusions           Follow links with probability α
                      Random jump with probability 1 − α
•••›››››››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting         Rewriting PageRank [Boldi et al., 2005]
PageRank

Functional                                            ∞
Rankings                                    (1 − α)
                                   r(α) =                   (αP)t .
Algorithms                                     N
                                                      t=0
Comparison

Conclusions
••••››››››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting
                  Definition (Branching contribution of a path)
PageRank
                  Given a path p = x1 , x2 , . . . , xt of length t = |p|
Functional
Rankings
                                                             1
Algorithms
                                    branching(p) =
Comparison                                            d1 d2 · · · dt−1
Conclusions       where di are the out-degrees of the members of the path
•••••›››››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                  Explicit formula for PageRank [Newman et al., 2001]
Notation
                                                   (1 − α)α|p|
Rewriting                 ri (α) =                             branching(p)
PageRank                                               N
                                     p∈Path(−,i)
Functional
Rankings

Algorithms

Comparison

Conclusions
•••••›››››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                  Explicit formula for PageRank [Newman et al., 2001]
Notation
                                                    (1 − α)α|p|
Rewriting                  ri (α) =                             branching(p)
PageRank                                                N
                                      p∈Path(−,i)
Functional
Rankings

Algorithms        Path(−, i) are incoming paths in node i
Comparison

Conclusions
•••••›››››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                  Explicit formula for PageRank [Newman et al., 2001]
Notation
                                                    (1 − α)α|p|
Rewriting                  ri (α) =                             branching(p)
PageRank                                                N
                                      p∈Path(−,i)
Functional
Rankings

Algorithms        Path(−, i) are incoming paths in node i
Comparison

Conclusions
                  General functional ranking
                                                   damping(|p|)
                          ri (α) =                              branching(p)
                                                       N
                                     p∈Path(−,i)
••••••››››››››››››
   Damping
 Functions for
 Link Ranking                                 Distribution of shortest paths
R. Baeza-Yates,
 P. Boldi and                  .it (40M pages)                                      .uk (18M pages)
   C. Castillo                0.3                                                  0.3

Notation
                              0.2                                                  0.2
                  Frequency




                                                                       Frequency
Rewriting
PageRank
                              0.1                                                  0.1
Functional
Rankings
                              0.0                                                  0.0
                                    5   10     15     20   25   30                       5   10     15     20   25   30
Algorithms
                                             Distance                                             Distance
Comparison
                     .eu.int (800K pages)                            Synthetic graph (100K pages)
Conclusions
                              0.3                                                  0.3


                              0.2                                                  0.2
                  Frequency




                                                                       Frequency
                              0.1                                                  0.1


                              0.0                                                  0.0
                                    5   10     15     20   25   30                       5   10     15     20   25   30
                                             Distance                                             Distance
•••••••›››››››››››
   Damping
 Functions for

                                0.30
 Link Ranking

R. Baeza-Yates,                                  damping(t) with α=0.8
 P. Boldi and
   C. Castillo
                                                 damping(t) with α=0.7

Notation                        0.20
Rewriting              Weight
PageRank

Functional
Rankings                        0.10
Algorithms

Comparison

Conclusions                     0.00
                                       1     2     3 4 5 6 7 8              9 10
                                                   Length of the path (t)

                  Exponential damping = PageRank
                                           damping(t) = α(1 − α)t
••••••••››››››››››
   Damping
 Functions for
                                 0.30
 Link Ranking
                                                damping(t) with L=15
R. Baeza-Yates,                                 damping(t) with L=10
 P. Boldi and
   C. Castillo
                                 0.20

                        Weight
Notation

Rewriting
PageRank

Functional                       0.10
Rankings

Algorithms

Comparison
                                 0.00
Conclusions                             1   2     3 4 5 6 7 8              9 10
                                                  Length of the path (t)

                  Linear damping
                                                          2(L−t)
                                                          L(L+1)   t<L
                                    damping(t) =
                                                         0         t≥L
••••••••››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                  For calculating LinearRank we use:
Notation
                                                   ∞
Rewriting                                      1
PageRank                     LinearRank =                damping(t)Pt
Functional                                     N
                                                   t=0
Rankings
                                                   L−1
Algorithms                                     1         2(L − t) t
Comparison
                                          =                       P
                                               N         L(L + 1)
Conclusions
                                                   t=0
••••••••››››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                  For calculating LinearRank we use:
Notation
                                                   ∞
Rewriting                                      1
PageRank                     LinearRank =                damping(t)Pt
Functional                                     N
                                                   t=0
Rankings
                                                   L−1
Algorithms                                     1         2(L − t) t
Comparison
                                          =                       P
                                               N         L(L + 1)
Conclusions
                                                   t=0


                  However, we cannot hold the temporary Pt in memory!
•••••••••›››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                  We have to rewrite to be able to calculate:
Notation
                                                  2
Rewriting                             R(0) =
PageRank                                        L+1
Functional                                      (L − k − 1) (k)
Rankings                           R(k+1) =                R P
Algorithms
                                                  (L − k)
Comparison

Conclusions
•••••••••›››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                  We have to rewrite to be able to calculate:
Notation
                                                  2
Rewriting                             R(0) =
PageRank                                        L+1
Functional                                      (L − k − 1) (k)
Rankings                           R(k+1) =                R P
Algorithms
                                                  (L − k)
Comparison
                                                L−1
Conclusions                    LinearRank =           R(k)
                                                k=0
•••••••••›››››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                  We have to rewrite to be able to calculate:
Notation
                                                   2
Rewriting                             R(0) =
PageRank                                         L+1
Functional                                       (L − k − 1) (k)
Rankings                           R(k+1) =                 R P
Algorithms
                                                   (L − k)
Comparison
                                                 L−1
Conclusions                    LinearRank =            R(k)
                                                k=0

                  Now we can give the algorithm . . .
••••••••••››››››››
   Damping
 Functions for
 Link Ranking     1: for i : 1 . . . N do {Initialization}
                                              2
R. Baeza-Yates,   2:   Score[i] ← R[i] ← L+1
 P. Boldi and
   C. Castillo    3: end for

Notation

Rewriting
PageRank

Functional
Rankings

Algorithms

Comparison

Conclusions
••••••••••››››››››
   Damping
 Functions for
 Link Ranking     1:   for i : 1 . . . N do {Initialization}
                                                2
R. Baeza-Yates,   2:     Score[i] ← R[i] ← L+1
 P. Boldi and
   C. Castillo    3:   end for
Notation
                  4:   for k : 1 . . . L − 1 do {Iteration step}
Rewriting
                  5:     Aux ← 0
PageRank

Functional
Rankings

Algorithms

Comparison

Conclusions
••••••••••››››››››
   Damping
 Functions for
 Link Ranking      1:   for i : 1 . . . N do {Initialization}
                                                 2
R. Baeza-Yates,    2:     Score[i] ← R[i] ← L+1
 P. Boldi and
   C. Castillo     3:   end for
Notation
                   4:   for k : 1 . . . L − 1 do {Iteration step}
Rewriting
                   5:     Aux ← 0
PageRank           6:     for i : 1 . . . N do {Follow links in the graph}
Functional
Rankings
                   7:         for all j such that there is a link from i to j do
Algorithms
                   8:           Aux[j] ← Aux[j] + R[i]/outdegree(i)
Comparison         9:         end for
Conclusions       10:     end for
••••••••••››››››››
   Damping
 Functions for
 Link Ranking      1:   for i : 1 . . . N do {Initialization}
                                                 2
R. Baeza-Yates,    2:     Score[i] ← R[i] ← L+1
 P. Boldi and
   C. Castillo     3:   end for
Notation
                   4:   for k : 1 . . . L − 1 do {Iteration step}
Rewriting
                   5:     Aux ← 0
PageRank           6:     for i : 1 . . . N do {Follow links in the graph}
Functional
Rankings
                   7:         for all j such that there is a link from i to j do
Algorithms
                   8:           Aux[j] ← Aux[j] + R[i]/outdegree(i)
Comparison         9:         end for
Conclusions       10:     end for
                  11:     for i : 1 . . . N do {Add to ranking value}
                  12:         R[i] ← Aux[i] × (L−k−1)
                                                 (L−k)
                  13:         Score[i] ← Score[i] + R[i]
                  14:     end for
                  15:   end for
                  16:   return Score
••••••••••››››››››
   Damping
 Functions for
 Link Ranking      1:   for i : 1 . . . N do {Initialization}
                                                 2
R. Baeza-Yates,    2:     Score[i] ← R[i] ← L+1
 P. Boldi and
   C. Castillo     3:   end for
Notation
                   4:   for k : 1 . . . L − 1 do {Iteration step}
Rewriting
                   5:     Aux ← 0
PageRank           6:     for i : 1 . . . N do {Follow links in the graph}
Functional
Rankings
                   7:         for all j such that there is a link from i to j do
Algorithms
                   8:           Aux[j] ← Aux[j] + R[i]/outdegree(i)
Comparison         9:         end for
Conclusions       10:     end for
                  11:     for i : 1 . . . N do {Add to ranking value}
                  12:         R[i] ← Aux[i] × (L−k−1)
                                                  (L−k)
                  13:         Score[i] ← Score[i] + R[i]
                  14:     end for
                  15:   end for
                  16:   return Score
•••••••••••›››››››
   Damping
 Functions for
 Link Ranking     Other functions studied in the paper:
R. Baeza-Yates,
 P. Boldi and         Hyperbolic damping
   C. Castillo

Notation

Rewriting
PageRank

Functional
Rankings

Algorithms

Comparison

Conclusions
•••••••••••›››››››
   Damping
 Functions for
 Link Ranking     Other functions studied in the paper:
R. Baeza-Yates,
 P. Boldi and         Hyperbolic damping
   C. Castillo
                      Empirical damping
Notation

Rewriting                                           0.7
PageRank                  Average text similarity
Functional
Rankings
                                                    0.6
Algorithms
                                                    0.5
Comparison

Conclusions
                                                    0.4

                                                    0.3

                                                    0.2
                                                          1   2         3       4   5
                                                                  Link distance
••••••••••••››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting
PageRank
                  How to approximate one functional ranking with another?
Functional
Rankings

Algorithms

Comparison

Conclusions
••••••••••••››››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting
PageRank
                  How to approximate one functional ranking with another?
Functional            Analysis (in the paper): match the first few levels of their
Rankings
                      damping functions
Algorithms

Comparison            In practice the orderings can be very similar . . .
Conclusions
•••••••••••••›››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation
                  Experimental comparison: 18-million nodes in the U.K. Web
Rewriting
PageRank          Graph
Functional
Rankings

Algorithms

Comparison

Conclusions
•••••••••••••›››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation
                  Experimental comparison: 18-million nodes in the U.K. Web
Rewriting
PageRank          Graph
Functional
Rankings
                      Calculated PageRank with α = 0.1, 0.2, . . . , 0.9
Algorithms

Comparison

Conclusions
•••••••••••••›››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation
                  Experimental comparison: 18-million nodes in the U.K. Web
Rewriting
PageRank          Graph
Functional
Rankings
                      Calculated PageRank with α = 0.1, 0.2, . . . , 0.9
Algorithms            Calculated LinearRank with L = 5, 10, . . . , 25
Comparison

Conclusions
•••••••••••••›››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation
                  Experimental comparison: 18-million nodes in the U.K. Web
Rewriting
PageRank          Graph
Functional
Rankings
                      Calculated PageRank with α = 0.1, 0.2, . . . , 0.9
Algorithms            Calculated LinearRank with L = 5, 10, . . . , 25
Comparison
                  For certain combinations of parameters, the rankings are
Conclusions
                  almost equal!
••••••••••••••››››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
                  Experimental Comparison in the U.K. Web Graph
   C. Castillo

Notation

Rewriting
                         1.00
                         0.95
                     τ
PageRank

Functional               0.90
Rankings
                         0.85
                                                                              τ ≥ 0.95
Algorithms               0.80
Comparison

Conclusions                 25
                                 20
                                      15                                      0.9
                                 L         10                 0.7
                                                                        0.8
                                                5 0.5   0.6         α
•••••••••••••••›››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
                  Prediction of Best Parameter Combinations (Analysis)
   C. Castillo
                                                     25
                                                                                  Actual optimum
Notation                                                          Predicted optimum with length=5
Rewriting
                      L that maximizes Kendall’s τ   20
PageRank

Functional
Rankings                                             15
Algorithms

Comparison                                           10
Conclusions

                                                      5


                                                          0.5   0.6         0.7             0.8     0.9
                                                                         Exponent α
••••••••••••••••››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting
                  What have we done?
PageRank

Functional
Rankings

Algorithms

Comparison

Conclusions
••••••••••••••••››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting
                  What have we done?
PageRank
                      Separate the damping from the calculation
Functional
Rankings

Algorithms

Comparison

Conclusions
••••••••••••••••››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting
                  What have we done?
PageRank
                      Separate the damping from the calculation
Functional
Rankings              Show that different damping functions can provide the
Algorithms
                      same ranking
Comparison

Conclusions
••••••••••••••••››
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting
                  What have we done?
PageRank
                      Separate the damping from the calculation
Functional
Rankings              Show that different damping functions can provide the
Algorithms
                      same ranking
Comparison

Conclusions
                      Analysis and experiments in the paper
•••••••••••••••••›
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting         What can we do with this?
PageRank

Functional            Fast approximation of PageRank using linear damping
Rankings

Algorithms

Comparison

Conclusions
•••••••••••••••••›
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting         What can we do with this?
PageRank

Functional            Fast approximation of PageRank using linear damping
Rankings

Algorithms
                      Fast calculation of other link-based rankings (e.g. HITS)
Comparison

Conclusions
•••••••••••••••••›
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting         What can we do with this?
PageRank

Functional            Fast approximation of PageRank using linear damping
Rankings

Algorithms
                      Fast calculation of other link-based rankings (e.g. HITS)
Comparison            Spam detection (e.g.: cut the first levels of links)
Conclusions
••••••••••••••••••
   Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo

Notation

Rewriting
PageRank

Functional
Rankings
                  Thank you!
Algorithms

Comparison

Conclusions
Damping
 Functions for
 Link Ranking

R. Baeza-Yates,
 P. Boldi and
   C. Castillo
                  Baeza-Yates, R., Boldi, P., and Castillo, C. (2005).
Notation          The choice of a damping function for propagating importance in link-based
                  ranking.
Rewriting
PageRank
                  Technical report, Dipartimento di Scienze dell’Informazione, Universit degli
                  Studi di Milano.
Functional
Rankings          Boldi, P., Santini, M., and Vigna, S. (2005).
Algorithms        Pagerank as a function of the damping factor.
                  In Proceedings of the 14th international conference on World Wide Web,
Comparison
                  pages 557–566, Chiba, Japan. ACM Press.
Conclusions
                  Newman, M. E., Strogatz, S. H., and Watts, D. J. (2001).
                  Random graphs with arbitrary degree distributions and their applications.
                  Phys Rev E Stat Nonlin Soft Matter Phys, 64(2 Pt 2).

Más contenido relacionado

Más de Carlos Castillo (ChaTo)

Finding High Quality Content in Social Media
Finding High Quality Content in Social MediaFinding High Quality Content in Social Media
Finding High Quality Content in Social MediaCarlos Castillo (ChaTo)
 
Socia Media and Digital Volunteering in Disaster Management @ DSEM 2017
Socia Media and Digital Volunteering in Disaster Management @ DSEM 2017Socia Media and Digital Volunteering in Disaster Management @ DSEM 2017
Socia Media and Digital Volunteering in Disaster Management @ DSEM 2017Carlos Castillo (ChaTo)
 
Detecting Algorithmic Bias (keynote at DIR 2016)
Detecting Algorithmic Bias (keynote at DIR 2016)Detecting Algorithmic Bias (keynote at DIR 2016)
Detecting Algorithmic Bias (keynote at DIR 2016)Carlos Castillo (ChaTo)
 

Más de Carlos Castillo (ChaTo) (20)

Finding High Quality Content in Social Media
Finding High Quality Content in Social MediaFinding High Quality Content in Social Media
Finding High Quality Content in Social Media
 
When no clicks are good news
When no clicks are good newsWhen no clicks are good news
When no clicks are good news
 
Socia Media and Digital Volunteering in Disaster Management @ DSEM 2017
Socia Media and Digital Volunteering in Disaster Management @ DSEM 2017Socia Media and Digital Volunteering in Disaster Management @ DSEM 2017
Socia Media and Digital Volunteering in Disaster Management @ DSEM 2017
 
Detecting Algorithmic Bias (keynote at DIR 2016)
Detecting Algorithmic Bias (keynote at DIR 2016)Detecting Algorithmic Bias (keynote at DIR 2016)
Detecting Algorithmic Bias (keynote at DIR 2016)
 
Discrimination Discovery
Discrimination DiscoveryDiscrimination Discovery
Discrimination Discovery
 
Fairness-Aware Data Mining
Fairness-Aware Data MiningFairness-Aware Data Mining
Fairness-Aware Data Mining
 
Big Crisis Data for ISPC
Big Crisis Data for ISPCBig Crisis Data for ISPC
Big Crisis Data for ISPC
 
Databeers: Big Crisis Data
Databeers: Big Crisis DataDatabeers: Big Crisis Data
Databeers: Big Crisis Data
 
Observational studies in social media
Observational studies in social mediaObservational studies in social media
Observational studies in social media
 
Natural experiments
Natural experimentsNatural experiments
Natural experiments
 
Content-based link prediction
Content-based link predictionContent-based link prediction
Content-based link prediction
 
Link prediction
Link predictionLink prediction
Link prediction
 
Recommender Systems
Recommender SystemsRecommender Systems
Recommender Systems
 
Graph Partitioning and Spectral Methods
Graph Partitioning and Spectral MethodsGraph Partitioning and Spectral Methods
Graph Partitioning and Spectral Methods
 
Finding Dense Subgraphs
Finding Dense SubgraphsFinding Dense Subgraphs
Finding Dense Subgraphs
 
Graph Evolution Models
Graph Evolution ModelsGraph Evolution Models
Graph Evolution Models
 
Link-Based Ranking
Link-Based RankingLink-Based Ranking
Link-Based Ranking
 
Text Indexing / Inverted Indices
Text Indexing / Inverted IndicesText Indexing / Inverted Indices
Text Indexing / Inverted Indices
 
Indexing
IndexingIndexing
Indexing
 
Text Summarization
Text SummarizationText Summarization
Text Summarization
 

Último

A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)Gabriella Davis
 
GenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdfGenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdflior mazor
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityPrincipled Technologies
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘RTylerCroy
 
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsMaria Levchenko
 
Apidays New York 2024 - The value of a flexible API Management solution for O...
Apidays New York 2024 - The value of a flexible API Management solution for O...Apidays New York 2024 - The value of a flexible API Management solution for O...
Apidays New York 2024 - The value of a flexible API Management solution for O...apidays
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024The Digital Insurer
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdfhans926745
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerThousandEyes
 
Real Time Object Detection Using Open CV
Real Time Object Detection Using Open CVReal Time Object Detection Using Open CV
Real Time Object Detection Using Open CVKhem
 
Strategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherStrategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherRemote DBA Services
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024The Digital Insurer
 
Artificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and MythsArtificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and MythsJoaquim Jorge
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc
 
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, AdobeApidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobeapidays
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationMichael W. Hawkins
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024The Digital Insurer
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024Rafal Los
 

Último (20)

A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
GenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdfGenAI Risks & Security Meetup 01052024.pdf
GenAI Risks & Security Meetup 01052024.pdf
 
Boost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivityBoost PC performance: How more available memory can improve productivity
Boost PC performance: How more available memory can improve productivity
 
🐬 The future of MySQL is Postgres 🐘
🐬  The future of MySQL is Postgres   🐘🐬  The future of MySQL is Postgres   🐘
🐬 The future of MySQL is Postgres 🐘
 
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data DiscoveryTrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
TrustArc Webinar - Unlock the Power of AI-Driven Data Discovery
 
Handwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed textsHandwritten Text Recognition for manuscripts and early printed texts
Handwritten Text Recognition for manuscripts and early printed texts
 
Apidays New York 2024 - The value of a flexible API Management solution for O...
Apidays New York 2024 - The value of a flexible API Management solution for O...Apidays New York 2024 - The value of a flexible API Management solution for O...
Apidays New York 2024 - The value of a flexible API Management solution for O...
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf
 
How to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected WorkerHow to Troubleshoot Apps for the Modern Connected Worker
How to Troubleshoot Apps for the Modern Connected Worker
 
Real Time Object Detection Using Open CV
Real Time Object Detection Using Open CVReal Time Object Detection Using Open CV
Real Time Object Detection Using Open CV
 
Strategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a FresherStrategies for Landing an Oracle DBA Job as a Fresher
Strategies for Landing an Oracle DBA Job as a Fresher
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024Partners Life - Insurer Innovation Award 2024
Partners Life - Insurer Innovation Award 2024
 
Artificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and MythsArtificial Intelligence: Facts and Myths
Artificial Intelligence: Facts and Myths
 
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law DevelopmentsTrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
TrustArc Webinar - Stay Ahead of US State Data Privacy Law Developments
 
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, AdobeApidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
Apidays New York 2024 - Scaling API-first by Ian Reasor and Radu Cotescu, Adobe
 
GenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day PresentationGenCyber Cyber Security Day Presentation
GenCyber Cyber Security Day Presentation
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024
 
The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024The 7 Things I Know About Cyber Security After 25 Years | April 2024
The 7 Things I Know About Cyber Security After 25 Years | April 2024
 

Generalizing PageRank (Pisa)

  • 1. Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo The Choice of a Damping Function for Notation Propagating Page Importance Rewriting PageRank in Link-Based Ranking Functional Rankings Algorithms Comparison Ricardo Baeza-Yates1 , Paolo Boldi2 and Carlos Castillo3 Conclusions 1. Yahoo Research – Barcelona, Spain 2. Universit` di Milano – Italy a 3. Universit` di Roma “La Sapienza” – Italy a February 6th, 2005
  • 2. Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and 1 Notation C. Castillo Notation 2 Rewriting PageRank Rewriting PageRank Functional Rankings 3 Functional Rankings Algorithms Comparison 4 Algorithms Conclusions 5 Comparison 6 Conclusions
  • 3. •››››››››››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting PageRank Let PN×N be the normalized link matrix of a graph Functional Rankings Row-normalized Algorithms No “sinks” Comparison Conclusions
  • 4. ••›››››››››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Definition (PageRank) Rewriting PageRank Stationary state of: Functional Rankings (1 − α) αP + 1N×N Algorithms N Comparison Conclusions
  • 5. ••›››››››››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Definition (PageRank) Rewriting PageRank Stationary state of: Functional Rankings (1 − α) αP + 1N×N Algorithms N Comparison Conclusions Follow links with probability α Random jump with probability 1 − α
  • 6. •••››››››››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting Rewriting PageRank [Boldi et al., 2005] PageRank Functional ∞ Rankings (1 − α) r(α) = (αP)t . Algorithms N t=0 Comparison Conclusions
  • 7. ••••›››››››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting Definition (Branching contribution of a path) PageRank Given a path p = x1 , x2 , . . . , xt of length t = |p| Functional Rankings 1 Algorithms branching(p) = Comparison d1 d2 · · · dt−1 Conclusions where di are the out-degrees of the members of the path
  • 8. •••••››››››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Explicit formula for PageRank [Newman et al., 2001] Notation (1 − α)α|p| Rewriting ri (α) = branching(p) PageRank N p∈Path(−,i) Functional Rankings Algorithms Comparison Conclusions
  • 9. •••••››››››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Explicit formula for PageRank [Newman et al., 2001] Notation (1 − α)α|p| Rewriting ri (α) = branching(p) PageRank N p∈Path(−,i) Functional Rankings Algorithms Path(−, i) are incoming paths in node i Comparison Conclusions
  • 10. •••••››››››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Explicit formula for PageRank [Newman et al., 2001] Notation (1 − α)α|p| Rewriting ri (α) = branching(p) PageRank N p∈Path(−,i) Functional Rankings Algorithms Path(−, i) are incoming paths in node i Comparison Conclusions General functional ranking damping(|p|) ri (α) = branching(p) N p∈Path(−,i)
  • 11. ••••••›››››››››››› Damping Functions for Link Ranking Distribution of shortest paths R. Baeza-Yates, P. Boldi and .it (40M pages) .uk (18M pages) C. Castillo 0.3 0.3 Notation 0.2 0.2 Frequency Frequency Rewriting PageRank 0.1 0.1 Functional Rankings 0.0 0.0 5 10 15 20 25 30 5 10 15 20 25 30 Algorithms Distance Distance Comparison .eu.int (800K pages) Synthetic graph (100K pages) Conclusions 0.3 0.3 0.2 0.2 Frequency Frequency 0.1 0.1 0.0 0.0 5 10 15 20 25 30 5 10 15 20 25 30 Distance Distance
  • 12. •••••••››››››››››› Damping Functions for 0.30 Link Ranking R. Baeza-Yates, damping(t) with α=0.8 P. Boldi and C. Castillo damping(t) with α=0.7 Notation 0.20 Rewriting Weight PageRank Functional Rankings 0.10 Algorithms Comparison Conclusions 0.00 1 2 3 4 5 6 7 8 9 10 Length of the path (t) Exponential damping = PageRank damping(t) = α(1 − α)t
  • 13. ••••••••›››››››››› Damping Functions for 0.30 Link Ranking damping(t) with L=15 R. Baeza-Yates, damping(t) with L=10 P. Boldi and C. Castillo 0.20 Weight Notation Rewriting PageRank Functional 0.10 Rankings Algorithms Comparison 0.00 Conclusions 1 2 3 4 5 6 7 8 9 10 Length of the path (t) Linear damping 2(L−t) L(L+1) t<L damping(t) = 0 t≥L
  • 14. ••••••••›››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo For calculating LinearRank we use: Notation ∞ Rewriting 1 PageRank LinearRank = damping(t)Pt Functional N t=0 Rankings L−1 Algorithms 1 2(L − t) t Comparison = P N L(L + 1) Conclusions t=0
  • 15. ••••••••›››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo For calculating LinearRank we use: Notation ∞ Rewriting 1 PageRank LinearRank = damping(t)Pt Functional N t=0 Rankings L−1 Algorithms 1 2(L − t) t Comparison = P N L(L + 1) Conclusions t=0 However, we cannot hold the temporary Pt in memory!
  • 16. •••••••••››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo We have to rewrite to be able to calculate: Notation 2 Rewriting R(0) = PageRank L+1 Functional (L − k − 1) (k) Rankings R(k+1) = R P Algorithms (L − k) Comparison Conclusions
  • 17. •••••••••››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo We have to rewrite to be able to calculate: Notation 2 Rewriting R(0) = PageRank L+1 Functional (L − k − 1) (k) Rankings R(k+1) = R P Algorithms (L − k) Comparison L−1 Conclusions LinearRank = R(k) k=0
  • 18. •••••••••››››››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo We have to rewrite to be able to calculate: Notation 2 Rewriting R(0) = PageRank L+1 Functional (L − k − 1) (k) Rankings R(k+1) = R P Algorithms (L − k) Comparison L−1 Conclusions LinearRank = R(k) k=0 Now we can give the algorithm . . .
  • 19. ••••••••••›››››››› Damping Functions for Link Ranking 1: for i : 1 . . . N do {Initialization} 2 R. Baeza-Yates, 2: Score[i] ← R[i] ← L+1 P. Boldi and C. Castillo 3: end for Notation Rewriting PageRank Functional Rankings Algorithms Comparison Conclusions
  • 20. ••••••••••›››››››› Damping Functions for Link Ranking 1: for i : 1 . . . N do {Initialization} 2 R. Baeza-Yates, 2: Score[i] ← R[i] ← L+1 P. Boldi and C. Castillo 3: end for Notation 4: for k : 1 . . . L − 1 do {Iteration step} Rewriting 5: Aux ← 0 PageRank Functional Rankings Algorithms Comparison Conclusions
  • 21. ••••••••••›››››››› Damping Functions for Link Ranking 1: for i : 1 . . . N do {Initialization} 2 R. Baeza-Yates, 2: Score[i] ← R[i] ← L+1 P. Boldi and C. Castillo 3: end for Notation 4: for k : 1 . . . L − 1 do {Iteration step} Rewriting 5: Aux ← 0 PageRank 6: for i : 1 . . . N do {Follow links in the graph} Functional Rankings 7: for all j such that there is a link from i to j do Algorithms 8: Aux[j] ← Aux[j] + R[i]/outdegree(i) Comparison 9: end for Conclusions 10: end for
  • 22. ••••••••••›››››››› Damping Functions for Link Ranking 1: for i : 1 . . . N do {Initialization} 2 R. Baeza-Yates, 2: Score[i] ← R[i] ← L+1 P. Boldi and C. Castillo 3: end for Notation 4: for k : 1 . . . L − 1 do {Iteration step} Rewriting 5: Aux ← 0 PageRank 6: for i : 1 . . . N do {Follow links in the graph} Functional Rankings 7: for all j such that there is a link from i to j do Algorithms 8: Aux[j] ← Aux[j] + R[i]/outdegree(i) Comparison 9: end for Conclusions 10: end for 11: for i : 1 . . . N do {Add to ranking value} 12: R[i] ← Aux[i] × (L−k−1) (L−k) 13: Score[i] ← Score[i] + R[i] 14: end for 15: end for 16: return Score
  • 23. ••••••••••›››››››› Damping Functions for Link Ranking 1: for i : 1 . . . N do {Initialization} 2 R. Baeza-Yates, 2: Score[i] ← R[i] ← L+1 P. Boldi and C. Castillo 3: end for Notation 4: for k : 1 . . . L − 1 do {Iteration step} Rewriting 5: Aux ← 0 PageRank 6: for i : 1 . . . N do {Follow links in the graph} Functional Rankings 7: for all j such that there is a link from i to j do Algorithms 8: Aux[j] ← Aux[j] + R[i]/outdegree(i) Comparison 9: end for Conclusions 10: end for 11: for i : 1 . . . N do {Add to ranking value} 12: R[i] ← Aux[i] × (L−k−1) (L−k) 13: Score[i] ← Score[i] + R[i] 14: end for 15: end for 16: return Score
  • 24. •••••••••••››››››› Damping Functions for Link Ranking Other functions studied in the paper: R. Baeza-Yates, P. Boldi and Hyperbolic damping C. Castillo Notation Rewriting PageRank Functional Rankings Algorithms Comparison Conclusions
  • 25. •••••••••••››››››› Damping Functions for Link Ranking Other functions studied in the paper: R. Baeza-Yates, P. Boldi and Hyperbolic damping C. Castillo Empirical damping Notation Rewriting 0.7 PageRank Average text similarity Functional Rankings 0.6 Algorithms 0.5 Comparison Conclusions 0.4 0.3 0.2 1 2 3 4 5 Link distance
  • 26. ••••••••••••›››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting PageRank How to approximate one functional ranking with another? Functional Rankings Algorithms Comparison Conclusions
  • 27. ••••••••••••›››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting PageRank How to approximate one functional ranking with another? Functional Analysis (in the paper): match the first few levels of their Rankings damping functions Algorithms Comparison In practice the orderings can be very similar . . . Conclusions
  • 28. •••••••••••••››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Experimental comparison: 18-million nodes in the U.K. Web Rewriting PageRank Graph Functional Rankings Algorithms Comparison Conclusions
  • 29. •••••••••••••››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Experimental comparison: 18-million nodes in the U.K. Web Rewriting PageRank Graph Functional Rankings Calculated PageRank with α = 0.1, 0.2, . . . , 0.9 Algorithms Comparison Conclusions
  • 30. •••••••••••••››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Experimental comparison: 18-million nodes in the U.K. Web Rewriting PageRank Graph Functional Rankings Calculated PageRank with α = 0.1, 0.2, . . . , 0.9 Algorithms Calculated LinearRank with L = 5, 10, . . . , 25 Comparison Conclusions
  • 31. •••••••••••••››››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Experimental comparison: 18-million nodes in the U.K. Web Rewriting PageRank Graph Functional Rankings Calculated PageRank with α = 0.1, 0.2, . . . , 0.9 Algorithms Calculated LinearRank with L = 5, 10, . . . , 25 Comparison For certain combinations of parameters, the rankings are Conclusions almost equal!
  • 32. ••••••••••••••›››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and Experimental Comparison in the U.K. Web Graph C. Castillo Notation Rewriting 1.00 0.95 τ PageRank Functional 0.90 Rankings 0.85 τ ≥ 0.95 Algorithms 0.80 Comparison Conclusions 25 20 15 0.9 L 10 0.7 0.8 5 0.5 0.6 α
  • 33. •••••••••••••••››› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and Prediction of Best Parameter Combinations (Analysis) C. Castillo 25 Actual optimum Notation Predicted optimum with length=5 Rewriting L that maximizes Kendall’s τ 20 PageRank Functional Rankings 15 Algorithms Comparison 10 Conclusions 5 0.5 0.6 0.7 0.8 0.9 Exponent α
  • 34. ••••••••••••••••›› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting What have we done? PageRank Functional Rankings Algorithms Comparison Conclusions
  • 35. ••••••••••••••••›› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting What have we done? PageRank Separate the damping from the calculation Functional Rankings Algorithms Comparison Conclusions
  • 36. ••••••••••••••••›› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting What have we done? PageRank Separate the damping from the calculation Functional Rankings Show that different damping functions can provide the Algorithms same ranking Comparison Conclusions
  • 37. ••••••••••••••••›› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting What have we done? PageRank Separate the damping from the calculation Functional Rankings Show that different damping functions can provide the Algorithms same ranking Comparison Conclusions Analysis and experiments in the paper
  • 38. •••••••••••••••••› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting What can we do with this? PageRank Functional Fast approximation of PageRank using linear damping Rankings Algorithms Comparison Conclusions
  • 39. •••••••••••••••••› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting What can we do with this? PageRank Functional Fast approximation of PageRank using linear damping Rankings Algorithms Fast calculation of other link-based rankings (e.g. HITS) Comparison Conclusions
  • 40. •••••••••••••••••› Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting What can we do with this? PageRank Functional Fast approximation of PageRank using linear damping Rankings Algorithms Fast calculation of other link-based rankings (e.g. HITS) Comparison Spam detection (e.g.: cut the first levels of links) Conclusions
  • 41. •••••••••••••••••• Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Notation Rewriting PageRank Functional Rankings Thank you! Algorithms Comparison Conclusions
  • 42. Damping Functions for Link Ranking R. Baeza-Yates, P. Boldi and C. Castillo Baeza-Yates, R., Boldi, P., and Castillo, C. (2005). Notation The choice of a damping function for propagating importance in link-based ranking. Rewriting PageRank Technical report, Dipartimento di Scienze dell’Informazione, Universit degli Studi di Milano. Functional Rankings Boldi, P., Santini, M., and Vigna, S. (2005). Algorithms Pagerank as a function of the damping factor. In Proceedings of the 14th international conference on World Wide Web, Comparison pages 557–566, Chiba, Japan. ACM Press. Conclusions Newman, M. E., Strogatz, S. H., and Watts, D. J. (2001). Random graphs with arbitrary degree distributions and their applications. Phys Rev E Stat Nonlin Soft Matter Phys, 64(2 Pt 2).