SlideShare a Scribd company logo
1 of 54
Download to read offline
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Recommender Systems for Social Tagging Systems
Leandro Balby Marinho
Machine Learning Lab
University of Hildesheim
PhD Defense
Leandro Balby Marinho 1 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Outline
1. Motivation
2. Problems and Contributions
3. Tag Recommender Systems
4. Nearest Neighbor-based Tag Recommendation
5. Cross-Tagging
6. Tag Enrichment
7. Conclusions and Future Work
Leandro Balby Marinho 2 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Web 2.0 sites more used than e-mail! [Nielsen Online (2009)]
In Web 2.0, the user plays the main role!
Leandro Balby Marinho 2 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Tags help users to organize and retrieve content.
Leandro Balby Marinho 3 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Tags also help other users to organize and retrieve their content.
Leandro Balby Marinho 4 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Folksonomy
A folksonomy is a structure F := (U, R, T, Y )
U ... users
R ... resources
T ... tags
Y ⊆ U × R × T ... tag assignments
X := {(u, r) | ∃t ∈ T : (u, r, t) ∈ Y } ... set of posts
Leandro Balby Marinho 5 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Outline
1. Motivation
2. Problems and Contributions
3. Tag Recommender Systems
4. Nearest Neighbor-based Tag Recommendation
5. Cross-Tagging
6. Tag Enrichment
7. Conclusions and Future Work
Leandro Balby Marinho 6 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Problems and Contributions
Tag Sparsity: Users are lazy to tag!
1 − |Y |
|U|×|R|×|T| ≈ 0.99 in all datasets used!
Solution: Tag Recommendation
Leandro Balby Marinho 6 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Problems and Contributions
Tag Sparsity: Users are lazy to tag!
1 − |Y |
|U|×|R|×|T| ≈ 0.99 in all datasets used!
Solution: Tag Recommendation
Social Network Divide: Compatible social systems are disconnected.
Leandro Balby Marinho 6 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Problems and Contributions
Tag Sparsity: Users are lazy to tag!
1 − |Y |
|U|×|R|×|T| ≈ 0.99 in all datasets used!
Solution: Tag Recommendation
Social Network Divide: Compatible social systems are disconnected.
Tag Idiosyncrasy: Tags bearing unclear semantics.
Solution: Tag Enrichment.
Leandro Balby Marinho 6 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Outline
1. Motivation
2. Problems and Contributions
3. Tag Recommender Systems
4. Nearest Neighbor-based Tag Recommendation
5. Cross-Tagging
6. Tag Enrichment
7. Conclusions and Future Work
Leandro Balby Marinho 7 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Tag Recommender Systems
...change the process from creation to recognition!
Personalized methods take the user preferences for tags into
consideration.
Value for the industry, e.g., youtube, flickr, last.fm, amazon.
Leandro Balby Marinho 7 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Evaluation and Metric
Xtrain ˙∪Xtest = X ... train/test splits based on posts
For each user, randomly pick one post for test.
Task: For (u, r) ∈ Xtest compute ˆT(u, r)
Metric: Recall((u, r) ∈ Xtest, n) := | ˆT(u,r)∩T(u,r)|
|T(u,r)|
Leandro Balby Marinho 8 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Formalization
Given (u, r) ∈ Xtest, a tag recommender system first computes:
Utility : {u} × {r} × T → R (1)
And then presents the tags in descending order of their utility:
ˆT(u, r) :=
n
argmax
t∈T
Utility(u, r, t) (2)
Leandro Balby Marinho 9 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Outline
1. Motivation
2. Problems and Contributions
3. Tag Recommender Systems
4. Nearest Neighbor-based Tag Recommendation
5. Cross-Tagging
6. Tag Enrichment
7. Conclusions and Future Work
Leandro Balby Marinho 10 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Nearest Neighbor-based (NN) Tag Recommenders
Collaborative Filtering (CF): Similar users tend to like similar things.
Here: Similar users tend to tag alike.
Traditional CF cannot be directly applied to folksonomies unless:
resources
tagsresources
users
users
userstags
Y
πUT YπURY
Leandro Balby Marinho 10 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Collaborative Filtering for Tag Recommendation
Neighborhood Formation: Nk
u :=
k
argmax
v∈Ur {u}
sim(mu, mv )
Recommendation:
ˆT(u, r) :=
n
argmax
t∈T
v∈Nk
u
sim(mu, mv )δ(v, r, t)
where δ(v, r, t) := 1 if (v, r, t) ∈ Y and 0 else.
Leandro Balby Marinho 11 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Ensembles of CF
Projections’ Ensemble:
Similarities’ Ensemble:
ˆT(u, r) =
n
argmax
t∈T
v∈Nu
(λsim(mu, mv ) + (1 − λ)sim(zu, zv ))δ(v, r, t)
where mu and mv are rows of πUT Y , and zu and zv rows of πUR Y .
Leandro Balby Marinho 12 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
A Graph-Based Tag Recommender based on Posts
We represent X as a homogeneous, undirected graph G := (X, E) over
the post set. Posts are related to each other if they share the same user:
Ruser := {(x, x ) ∈ X × X | user(x) = user(x )}
the same resource:
Rres := {(x, x ) ∈ X × X|res(x) = res(x )}
or either share the same user or resource:
Rres
user := Ruser ∪ Rres
where user(x) and res(x) are the user and resource associated with the
post x respectively.
Leandro Balby Marinho 13 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Relational Graph based on Posts
Leandro Balby Marinho 14 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Weighting Schemes
For x ∈ Xtest and (x, x ) ∈ E:
1. User-Tag Profile:
φuser-tag
:= (|Y ∩ ({user(x)} × R × {t})|)t∈T
2. Resource-Tag Profile:
φres-tag
:= (|Y ∩ (U × {res(x)} × {t})|)t∈T
Weight:
w(x, x ) :=
φ(x), φ(x )
φ(x) φ(x )
Leandro Balby Marinho 15 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Relational Classification
Weighted Average (WA) [Marinho et al. (2009)]:
P(t|x) :=
x ∈Nx |t∈T(x ) w(x, x )
x ∈Nx
w(x, x )
where:
Nx := {x ∈ X | (x, x ) ∈ R, T(x) = ∅}
Runtime: O (|T||Nx |))
Leandro Balby Marinho 16 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Evaluation
Datasets:
dataset |U| |R| |T| Triples |Y | Posts |X|
BibSonomy 116 361 412 10,148 2,522
Last.fm 2,917 1,853 2,045 219,702 75,565
Delicious 37,399 74,874 22,170 7,487,319 3,055,436
Evaluated methods:
Baselines: (Locally) Constant Models (GCT,LCR, LCU).
Ensemble of Locally Constant Models (LCE) [J¨aschke et al. 2008].
TopicRank, FolkRank [J¨aschke et al. 2007]
RTF [Rendle et al. 2009]
PITF [Rendle et al. 2010]
Our NN-based Recommenders
Leandro Balby Marinho 17 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Results: NN Methods
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10
Recall
Number of recommended tags
Top-10 Tag Recommendations in Delicious
WA
CF UT
CF UR
matrixExt
simEns
LCR
GCT
Leandro Balby Marinho 18 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Results: WA vs. State-of-the-Art
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10
Recall
Number of recommended tags
Top-10 Tag Recommendations in BibSonomy
WA
RTF
PITF
FolkRank
LCE
TopicRank
Leandro Balby Marinho 19 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Results: WA vs. State-of-the-Art
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10
Recall
Number of recommended tags
Top-10 Tag Recommendations in Last.fm
PITF
WA
RTF
FolkRank
LCE
TopicRank
Leandro Balby Marinho 20 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Results: WA vs. State-of-the-Art
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10
Recall
Number of recommended tags
Top-10 Tag Recommendations in Delicious
PITF
WA
FolkRank
LCE
TopicRank
Leandro Balby Marinho 21 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Runtime: WA vs. PITF
BibSonomy Last.fm Delicious
Method Runtime Runtime Runtime
WA < 1 second < 1 minute ≈ 3 minutes
PITF ≈ 5 minutes ≈ 7 hours ≈ 33 days
Leandro Balby Marinho 22 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
ECML/Discovery Challenge 2009
2nd Place ECML/PKDD Discovery Challenge 2009!
Rank Method Top-5 F1
1 PITF [Rendle et al. (2009)] 0.35594
2 Relational Ensemble [Marinho et al. (2009)]1
0.33185
– WA (not submitted) 0.32519
3 Content-based [Lipczak et al. (2009)] 0.32461
1
With Christine Preisach
Leandro Balby Marinho 23 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Outline
1. Motivation
2. Problems and Contributions
3. Tag Recommender Systems
4. Nearest Neighbor-based Tag Recommendation
5. Cross-Tagging
6. Tag Enrichment
7. Conclusions and Future Work
Leandro Balby Marinho 24 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Problem
Use resources overlap to cross tags between systems.
Leandro Balby Marinho 24 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Tag Recommendation for Cross-Tagging
Cross-Tagging Approaches:
LCR (locally constant per resource).
Collaborative Filtering.
Leandro Balby Marinho 25 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Evaluation
Tag-Aware-based Evaluation
The better the tags the better a tag-aware recommender that uses
those tags.
Tag-Aware based on HOSVD [Symeonidis et al. (2008)]
Datasets
Blogger.com Last.fm Annotated Blog
|U| 6,620 44,143 3,827
|R| 17,372 17,372 1,323
|T| 0 4,903 422
|Y | 0 254,388 32,900
Leandro Balby Marinho 26 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Recall on the top-5 resources of HOSVD
n - Number of tags used to annotate the test posts of Blogger.com.
Leandro Balby Marinho 27 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Outline
1. Motivation
2. Problems and Contributions
3. Tag Recommender Systems
4. Nearest Neighbor-based Tag Recommendation
5. Cross-Tagging
6. Tag Enrichment
7. Conclusions and Future Work
Leandro Balby Marinho 28 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Problems
First we map tags from a folksonomy to concepts C of an ontology
H : T → C
Then we learn an ontology P such that:
CP := T ˙∪ C
The better the ontology the better a ontology-aware recommender
that uses this ontology.
Taxonomy driven CF [Ziegler et al. (2004)]
Datasets:
dataset |U| |T| |R| |Y |
Last.fm 3,532 7,081 982 130,899
musicmoz - 555 982 -
Leandro Balby Marinho 28 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Results
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Trivial Ontology Domain Expert Ontology Learned Ontology
Recall
Leandro Balby Marinho 29 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Outline
1. Motivation
2. Problems and Contributions
3. Tag Recommender Systems
4. Nearest Neighbor-based Tag Recommendation
5. Cross-Tagging
6. Tag Enrichment
7. Conclusions and Future Work
Leandro Balby Marinho 30 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Conclusions
Tag Sparsity: Nearest Neighbor Method that
Performs competitively to more sophisticated methods.
Require modest computational effort.
Social Network Divide:
Cross-tagging as a tag recommendation problem.
Personalized cross-tagging better than non-personalized
cross-tagging.
Tag idiosyncrasy: Tag enrichment
Well agreed concepts that match the semantic intention of
users.
Learned ontology better than trivial or domain expert ontology.
New recommender systems-based evaluation protocols.
Leandro Balby Marinho 30 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Future Work
Optimzed weight learning for WA.
Bidirectional Cross-Tagging.
Optimized Cross-Tagging/Ontology learning.
Leandro Balby Marinho 31 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Results NN vs. Baselines
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10
Recall
Number of recommended tags
Top-10 Tag Recommendations in BibSonomy
WA
CF UT
CF UR
matrixExt
simEns
LCR
GCT
0
0.2
0.4
0.6
0.8
1
0 2 4 6 8 10Recall
Number of recommended tags
Top-10 Tag Recommendations in Last.fm
WA
CF UT
CF UR
matrixExt
simEns
LCR
GCT
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
PageRank for Folksonomias
Based on PageRank [Hotho et al. 2006]
Each hyperedge is broken into three undirected edges:
Now PageRank can be applied:
wt+1 ← λAT
wt + (1 − λ)p
Rank will be dominated by popular nodes (Skewd distribution of tag
assignments)
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
FolkRank
1. First compute vector w(0)
with p = 1.
2. Next compute vector w(1)
with p[u] := 1 + |U|, p[r] := 1 + |R|, and
p[v] := 1 for v = u, r.
3. Finally compute w := w(1)
− w(0)
.
4. Recommendation list ˆT(u, r) is the top-n nodes in the rank
restricted to tags.
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
RTF: Ranking with Tensor Factorization
Tag Recommendation as a tensor completion problem.
Positive tags have higher rank than negative ones [Rendle et al. 2009].
yu,r,t1 > yu,r,t2 ⇔ (u, r, t1) ∈ T+
u,r ∧ (u, r, t2) ∈ T−
u,r
T+
u,r := {t | (u, r) ∈ Xtreino ∧ (u, r, t) ∈ Y }, T−
u,r := {t | (u, r) ∈ Xtreino ∧ (u, r, t)
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Tucker Decomposition Model
ˆY := ˆC ×u
ˆU ×r
ˆR ×t
ˆT
or equivalently:
ˆyu,r,t =
˜u ˜r ˜t
ˆc˜u,˜r,˜t · ˆuu,˜u · ˆrr,˜r · ˆtt,˜t
where the model parameters are:
ˆC ∈ RkU ×kR ×kT
, ˆU ∈ R|U|×kU
, ˆR ∈ R|R|×kR
, ˆT ∈ R|T|×kT
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
PITF: Pairwise Interaction Tensor Factorization
PITF only models the two-way interactions between user and tags as well
as between resources and tags:
ˆau,r,t =
k
f
ˆuu,f · ˆtU
t,f +
k
f
ˆrr,f · ˆtR
t,f
where ˆU ∈ R|U|×k
, ˆR ∈ R|R|×k
, ˆTU
∈ R|T|×k
and ˆTR
∈ R|T|×k
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Complexity
Learning Runtime Complexity
Method Runtime
WA O(1)
FolkRank O(1)
RTF O iter · |Xtrain||T|2
· kU · kR · kT
PITF O(iter · |Xtrain||T|2
· 2k)
Prediction Runtime Complexity
Method Runtime
WA O (|T||Nx | + |T| log(n)))
Folkrank O(iter · (|Y | + |U| + |R| + |T|) + |T| + |T| log(n))
RTF O(|T| · kU + kR · kT · kT )
PITF O(|T|2k + |T| log(n))
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Relation Rewarding
We can reward the best relation by a factor c ∈ R
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Results Cross-Tagging
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Tag Enrichment Approach
Semantic mapping as an ontology matching problem.
P(A, B) ≈ | A ∩ B |
|R| [Doan et al. (2004)]
Jaccard coefficient:
JS(A, B) := P(A ∩ B)/P(A ∪ B) :=
P(A, B)
P(A, B) + P(A, ¯B) + P(¯A, B)
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Ontology learning
Frequent itemset mining for ontology learning [Marinho et al. 2008]2
.
2
Algorithm proposed by Krisztian Buza co-author of [Marinho et al. 2008]
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
Semantic mapping
tags mapped concepts
electro electronica
hip hop hip hop
chillout rock
old skool dance house
anything else but death heavy metal
post-hardcore emo
california punk
political punk
urban hip hop
60s stuff country-rock
relaxing folk rock
explorer experimental rock
rock en espanol latin pop
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
An Extract of Domain Expert Ontology
heavy_metal
death_metal
doom_metal
black_metal
thrash rap-metal
hair_metal
speed_metal
grindcore
metal
Pajek
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions
An Extract of Learned Ontology
maynard james keenan
powerful
technical death metal
brit-rock
metalcore
nu metal
doom metal
new age
finnishprogressive metal
alternative metal
melodic death metal
melancholic
black metal
progressive
ethereal
swedish metal
gothenburg metal
progressive death metal
german
bands i have seen live
speed metal
nwobhm
heavy
power metal
symphonic metal
guitargasm
death-doom metal
gothic metal
famous frontman
art rock
viking metal
groove metal
melodic metal
violent
aggressive alternative - at work music
moody
faves
a-o-t-w
slipknot
grindcore
great lyrics
gothenburg
dark
g00ds
70s progressive rock
depressing
cold
doom
art-rock
prog
trash metal
depression
brutal death metal
us
loud
sad
korn
soad
mezmerize
fall out boy
rap-metal
seen them live
nu-metal
cello metal
melodic black metal
folk metal
guitar music
symphonic prog
british metal
awesome
zeuhl
female fronted metal
love metal
aggressive
finland
epic
nellis1
symphonic black metal
new metal
ominous
buen metal
bands ive seen live
classic thrash
bands i have seen
prog metal
classic metal
prog rock metal gods
my band inspiration
metal of some persuasionfavorite shitnice music
grooving metal
fav artistsblizzards main tags
symphonic death
grind
melodic power metal
everything
speed
favs
melodic death
heavy_metal
death
progressive_rock
doom_metal
death_metal
thrash
metal
speed_metal
periods
Pajek
Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim

More Related Content

Similar to Tag Recommender Systems Using Nearest Neighbors and Graph-Based Methods

(Machine)Learning with limited labels(Machine)Learning with limited labels(Ma...
(Machine)Learning with limited labels(Machine)Learning with limited labels(Ma...(Machine)Learning with limited labels(Machine)Learning with limited labels(Ma...
(Machine)Learning with limited labels(Machine)Learning with limited labels(Ma...Eirini Ntoutsi
 
Learning Content and Usage Factors Simultaneously
Learning Content and Usage Factors SimultaneouslyLearning Content and Usage Factors Simultaneously
Learning Content and Usage Factors SimultaneouslyArnab Bhadury
 
Data.Mining.C.6(II).classification and prediction
Data.Mining.C.6(II).classification and predictionData.Mining.C.6(II).classification and prediction
Data.Mining.C.6(II).classification and predictionMargaret Wang
 
Machine learning - session 3
Machine learning - session 3Machine learning - session 3
Machine learning - session 3Luis Borbon
 
SPSS statistics - get help using SPSS
SPSS statistics - get help using SPSSSPSS statistics - get help using SPSS
SPSS statistics - get help using SPSScsula its training
 
lecture_mooney.ppt
lecture_mooney.pptlecture_mooney.ppt
lecture_mooney.pptbutest
 
Machine Learning and Artificial Neural Networks.ppt
Machine Learning and Artificial Neural Networks.pptMachine Learning and Artificial Neural Networks.ppt
Machine Learning and Artificial Neural Networks.pptAnshika865276
 
An Introduction to SPSS
An Introduction to SPSSAn Introduction to SPSS
An Introduction to SPSSRayman Soe
 
Machine Learning for NLP
Machine Learning for NLPMachine Learning for NLP
Machine Learning for NLPbutest
 
Control of Photo Sharing on Online Social Network.
Control of Photo Sharing on Online Social Network.Control of Photo Sharing on Online Social Network.
Control of Photo Sharing on Online Social Network.SAFAD ISMAIL
 
A Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social MediaA Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social MediaEditor IJCATR
 
A Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social MediaA Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social MediaEditor IJCATR
 
A Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social MediaA Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social MediaEditor IJCATR
 

Similar to Tag Recommender Systems Using Nearest Neighbors and Graph-Based Methods (20)

Introduction to spss
Introduction to spssIntroduction to spss
Introduction to spss
 
(Machine)Learning with limited labels(Machine)Learning with limited labels(Ma...
(Machine)Learning with limited labels(Machine)Learning with limited labels(Ma...(Machine)Learning with limited labels(Machine)Learning with limited labels(Ma...
(Machine)Learning with limited labels(Machine)Learning with limited labels(Ma...
 
Ankit presentation
Ankit presentationAnkit presentation
Ankit presentation
 
Learning Content and Usage Factors Simultaneously
Learning Content and Usage Factors SimultaneouslyLearning Content and Usage Factors Simultaneously
Learning Content and Usage Factors Simultaneously
 
Data.Mining.C.6(II).classification and prediction
Data.Mining.C.6(II).classification and predictionData.Mining.C.6(II).classification and prediction
Data.Mining.C.6(II).classification and prediction
 
CSC410-Presentation
CSC410-PresentationCSC410-Presentation
CSC410-Presentation
 
Classroom Research
Classroom ResearchClassroom Research
Classroom Research
 
Machine learning - session 3
Machine learning - session 3Machine learning - session 3
Machine learning - session 3
 
Joseph Jay Williams - WESST - Bridging Research and Practice via MOOClets & C...
Joseph Jay Williams - WESST - Bridging Research and Practice via MOOClets & C...Joseph Jay Williams - WESST - Bridging Research and Practice via MOOClets & C...
Joseph Jay Williams - WESST - Bridging Research and Practice via MOOClets & C...
 
SPSS statistics - get help using SPSS
SPSS statistics - get help using SPSSSPSS statistics - get help using SPSS
SPSS statistics - get help using SPSS
 
lecture_mooney.ppt
lecture_mooney.pptlecture_mooney.ppt
lecture_mooney.ppt
 
Machine Learning and Artificial Neural Networks.ppt
Machine Learning and Artificial Neural Networks.pptMachine Learning and Artificial Neural Networks.ppt
Machine Learning and Artificial Neural Networks.ppt
 
An Introduction to SPSS
An Introduction to SPSSAn Introduction to SPSS
An Introduction to SPSS
 
Machine Learning for NLP
Machine Learning for NLPMachine Learning for NLP
Machine Learning for NLP
 
Joseph Jay Williams - WESST - Bridging Research via MOOClets and Collaborativ...
Joseph Jay Williams - WESST - Bridging Research via MOOClets and Collaborativ...Joseph Jay Williams - WESST - Bridging Research via MOOClets and Collaborativ...
Joseph Jay Williams - WESST - Bridging Research via MOOClets and Collaborativ...
 
Control of Photo Sharing on Online Social Network.
Control of Photo Sharing on Online Social Network.Control of Photo Sharing on Online Social Network.
Control of Photo Sharing on Online Social Network.
 
A Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social MediaA Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social Media
 
A Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social MediaA Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social Media
 
A Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social MediaA Survey on Decision Support Systems in Social Media
A Survey on Decision Support Systems in Social Media
 
master_thesis.pdf
master_thesis.pdfmaster_thesis.pdf
master_thesis.pdf
 

Recently uploaded

Nutritional Needs Presentation - HLTH 104
Nutritional Needs Presentation - HLTH 104Nutritional Needs Presentation - HLTH 104
Nutritional Needs Presentation - HLTH 104misteraugie
 
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...fonyou31
 
Paris 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activityParis 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activityGeoBlogs
 
The Most Excellent Way | 1 Corinthians 13
The Most Excellent Way | 1 Corinthians 13The Most Excellent Way | 1 Corinthians 13
The Most Excellent Way | 1 Corinthians 13Steve Thomason
 
microwave assisted reaction. General introduction
microwave assisted reaction. General introductionmicrowave assisted reaction. General introduction
microwave assisted reaction. General introductionMaksud Ahmed
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfsanyamsingh5019
 
Russian Call Girls in Andheri Airport Mumbai WhatsApp 9167673311 💞 Full Nigh...
Russian Call Girls in Andheri Airport Mumbai WhatsApp  9167673311 💞 Full Nigh...Russian Call Girls in Andheri Airport Mumbai WhatsApp  9167673311 💞 Full Nigh...
Russian Call Girls in Andheri Airport Mumbai WhatsApp 9167673311 💞 Full Nigh...Pooja Nehwal
 
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxiammrhaywood
 
The basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptxThe basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptxheathfieldcps1
 
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Sapana Sha
 
9548086042 for call girls in Indira Nagar with room service
9548086042  for call girls in Indira Nagar  with room service9548086042  for call girls in Indira Nagar  with room service
9548086042 for call girls in Indira Nagar with room servicediscovermytutordmt
 
social pharmacy d-pharm 1st year by Pragati K. Mahajan
social pharmacy d-pharm 1st year by Pragati K. Mahajansocial pharmacy d-pharm 1st year by Pragati K. Mahajan
social pharmacy d-pharm 1st year by Pragati K. Mahajanpragatimahajan3
 
CARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxCARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxGaneshChakor2
 
Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)eniolaolutunde
 
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions  for the students and aspirants of Chemistry12th.pptxOrganic Name Reactions  for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions for the students and aspirants of Chemistry12th.pptxVS Mahajan Coaching Centre
 
1029-Danh muc Sach Giao Khoa khoi 6.pdf
1029-Danh muc Sach Giao Khoa khoi  6.pdf1029-Danh muc Sach Giao Khoa khoi  6.pdf
1029-Danh muc Sach Giao Khoa khoi 6.pdfQucHHunhnh
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAssociation for Project Management
 
Mastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionMastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionSafetyChain Software
 
Grant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingGrant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingTechSoup
 

Recently uploaded (20)

Nutritional Needs Presentation - HLTH 104
Nutritional Needs Presentation - HLTH 104Nutritional Needs Presentation - HLTH 104
Nutritional Needs Presentation - HLTH 104
 
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
Ecosystem Interactions Class Discussion Presentation in Blue Green Lined Styl...
 
Paris 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activityParis 2024 Olympic Geographies - an activity
Paris 2024 Olympic Geographies - an activity
 
The Most Excellent Way | 1 Corinthians 13
The Most Excellent Way | 1 Corinthians 13The Most Excellent Way | 1 Corinthians 13
The Most Excellent Way | 1 Corinthians 13
 
microwave assisted reaction. General introduction
microwave assisted reaction. General introductionmicrowave assisted reaction. General introduction
microwave assisted reaction. General introduction
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdf
 
Advance Mobile Application Development class 07
Advance Mobile Application Development class 07Advance Mobile Application Development class 07
Advance Mobile Application Development class 07
 
Russian Call Girls in Andheri Airport Mumbai WhatsApp 9167673311 💞 Full Nigh...
Russian Call Girls in Andheri Airport Mumbai WhatsApp  9167673311 💞 Full Nigh...Russian Call Girls in Andheri Airport Mumbai WhatsApp  9167673311 💞 Full Nigh...
Russian Call Girls in Andheri Airport Mumbai WhatsApp 9167673311 💞 Full Nigh...
 
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptxSOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
SOCIAL AND HISTORICAL CONTEXT - LFTVD.pptx
 
The basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptxThe basics of sentences session 2pptx copy.pptx
The basics of sentences session 2pptx copy.pptx
 
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
 
9548086042 for call girls in Indira Nagar with room service
9548086042  for call girls in Indira Nagar  with room service9548086042  for call girls in Indira Nagar  with room service
9548086042 for call girls in Indira Nagar with room service
 
social pharmacy d-pharm 1st year by Pragati K. Mahajan
social pharmacy d-pharm 1st year by Pragati K. Mahajansocial pharmacy d-pharm 1st year by Pragati K. Mahajan
social pharmacy d-pharm 1st year by Pragati K. Mahajan
 
CARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxCARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptx
 
Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)Software Engineering Methodologies (overview)
Software Engineering Methodologies (overview)
 
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions  for the students and aspirants of Chemistry12th.pptxOrganic Name Reactions  for the students and aspirants of Chemistry12th.pptx
Organic Name Reactions for the students and aspirants of Chemistry12th.pptx
 
1029-Danh muc Sach Giao Khoa khoi 6.pdf
1029-Danh muc Sach Giao Khoa khoi  6.pdf1029-Danh muc Sach Giao Khoa khoi  6.pdf
1029-Danh muc Sach Giao Khoa khoi 6.pdf
 
APM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across SectorsAPM Welcome, APM North West Network Conference, Synergies Across Sectors
APM Welcome, APM North West Network Conference, Synergies Across Sectors
 
Mastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory InspectionMastering the Unannounced Regulatory Inspection
Mastering the Unannounced Regulatory Inspection
 
Grant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy ConsultingGrant Readiness 101 TechSoup and Remy Consulting
Grant Readiness 101 TechSoup and Remy Consulting
 

Tag Recommender Systems Using Nearest Neighbors and Graph-Based Methods

  • 1. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Recommender Systems for Social Tagging Systems Leandro Balby Marinho Machine Learning Lab University of Hildesheim PhD Defense Leandro Balby Marinho 1 / 32 Machine Learning Lab, University of Hildesheim
  • 2. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Outline 1. Motivation 2. Problems and Contributions 3. Tag Recommender Systems 4. Nearest Neighbor-based Tag Recommendation 5. Cross-Tagging 6. Tag Enrichment 7. Conclusions and Future Work Leandro Balby Marinho 2 / 32 Machine Learning Lab, University of Hildesheim
  • 3. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Web 2.0 sites more used than e-mail! [Nielsen Online (2009)] In Web 2.0, the user plays the main role! Leandro Balby Marinho 2 / 32 Machine Learning Lab, University of Hildesheim
  • 4. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Tags help users to organize and retrieve content. Leandro Balby Marinho 3 / 32 Machine Learning Lab, University of Hildesheim
  • 5. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Tags also help other users to organize and retrieve their content. Leandro Balby Marinho 4 / 32 Machine Learning Lab, University of Hildesheim
  • 6. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Folksonomy A folksonomy is a structure F := (U, R, T, Y ) U ... users R ... resources T ... tags Y ⊆ U × R × T ... tag assignments X := {(u, r) | ∃t ∈ T : (u, r, t) ∈ Y } ... set of posts Leandro Balby Marinho 5 / 32 Machine Learning Lab, University of Hildesheim
  • 7. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Outline 1. Motivation 2. Problems and Contributions 3. Tag Recommender Systems 4. Nearest Neighbor-based Tag Recommendation 5. Cross-Tagging 6. Tag Enrichment 7. Conclusions and Future Work Leandro Balby Marinho 6 / 32 Machine Learning Lab, University of Hildesheim
  • 8. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Problems and Contributions Tag Sparsity: Users are lazy to tag! 1 − |Y | |U|×|R|×|T| ≈ 0.99 in all datasets used! Solution: Tag Recommendation Leandro Balby Marinho 6 / 32 Machine Learning Lab, University of Hildesheim
  • 9. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Problems and Contributions Tag Sparsity: Users are lazy to tag! 1 − |Y | |U|×|R|×|T| ≈ 0.99 in all datasets used! Solution: Tag Recommendation Social Network Divide: Compatible social systems are disconnected. Leandro Balby Marinho 6 / 32 Machine Learning Lab, University of Hildesheim
  • 10. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Problems and Contributions Tag Sparsity: Users are lazy to tag! 1 − |Y | |U|×|R|×|T| ≈ 0.99 in all datasets used! Solution: Tag Recommendation Social Network Divide: Compatible social systems are disconnected. Tag Idiosyncrasy: Tags bearing unclear semantics. Solution: Tag Enrichment. Leandro Balby Marinho 6 / 32 Machine Learning Lab, University of Hildesheim
  • 11. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Outline 1. Motivation 2. Problems and Contributions 3. Tag Recommender Systems 4. Nearest Neighbor-based Tag Recommendation 5. Cross-Tagging 6. Tag Enrichment 7. Conclusions and Future Work Leandro Balby Marinho 7 / 32 Machine Learning Lab, University of Hildesheim
  • 12. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Tag Recommender Systems ...change the process from creation to recognition! Personalized methods take the user preferences for tags into consideration. Value for the industry, e.g., youtube, flickr, last.fm, amazon. Leandro Balby Marinho 7 / 32 Machine Learning Lab, University of Hildesheim
  • 13. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Evaluation and Metric Xtrain ˙∪Xtest = X ... train/test splits based on posts For each user, randomly pick one post for test. Task: For (u, r) ∈ Xtest compute ˆT(u, r) Metric: Recall((u, r) ∈ Xtest, n) := | ˆT(u,r)∩T(u,r)| |T(u,r)| Leandro Balby Marinho 8 / 32 Machine Learning Lab, University of Hildesheim
  • 14. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Formalization Given (u, r) ∈ Xtest, a tag recommender system first computes: Utility : {u} × {r} × T → R (1) And then presents the tags in descending order of their utility: ˆT(u, r) := n argmax t∈T Utility(u, r, t) (2) Leandro Balby Marinho 9 / 32 Machine Learning Lab, University of Hildesheim
  • 15. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Outline 1. Motivation 2. Problems and Contributions 3. Tag Recommender Systems 4. Nearest Neighbor-based Tag Recommendation 5. Cross-Tagging 6. Tag Enrichment 7. Conclusions and Future Work Leandro Balby Marinho 10 / 32 Machine Learning Lab, University of Hildesheim
  • 16. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Nearest Neighbor-based (NN) Tag Recommenders Collaborative Filtering (CF): Similar users tend to like similar things. Here: Similar users tend to tag alike. Traditional CF cannot be directly applied to folksonomies unless: resources tagsresources users users userstags Y πUT YπURY Leandro Balby Marinho 10 / 32 Machine Learning Lab, University of Hildesheim
  • 17. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Collaborative Filtering for Tag Recommendation Neighborhood Formation: Nk u := k argmax v∈Ur {u} sim(mu, mv ) Recommendation: ˆT(u, r) := n argmax t∈T v∈Nk u sim(mu, mv )δ(v, r, t) where δ(v, r, t) := 1 if (v, r, t) ∈ Y and 0 else. Leandro Balby Marinho 11 / 32 Machine Learning Lab, University of Hildesheim
  • 18. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Ensembles of CF Projections’ Ensemble: Similarities’ Ensemble: ˆT(u, r) = n argmax t∈T v∈Nu (λsim(mu, mv ) + (1 − λ)sim(zu, zv ))δ(v, r, t) where mu and mv are rows of πUT Y , and zu and zv rows of πUR Y . Leandro Balby Marinho 12 / 32 Machine Learning Lab, University of Hildesheim
  • 19. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions A Graph-Based Tag Recommender based on Posts We represent X as a homogeneous, undirected graph G := (X, E) over the post set. Posts are related to each other if they share the same user: Ruser := {(x, x ) ∈ X × X | user(x) = user(x )} the same resource: Rres := {(x, x ) ∈ X × X|res(x) = res(x )} or either share the same user or resource: Rres user := Ruser ∪ Rres where user(x) and res(x) are the user and resource associated with the post x respectively. Leandro Balby Marinho 13 / 32 Machine Learning Lab, University of Hildesheim
  • 20. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Relational Graph based on Posts Leandro Balby Marinho 14 / 32 Machine Learning Lab, University of Hildesheim
  • 21. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Weighting Schemes For x ∈ Xtest and (x, x ) ∈ E: 1. User-Tag Profile: φuser-tag := (|Y ∩ ({user(x)} × R × {t})|)t∈T 2. Resource-Tag Profile: φres-tag := (|Y ∩ (U × {res(x)} × {t})|)t∈T Weight: w(x, x ) := φ(x), φ(x ) φ(x) φ(x ) Leandro Balby Marinho 15 / 32 Machine Learning Lab, University of Hildesheim
  • 22. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Relational Classification Weighted Average (WA) [Marinho et al. (2009)]: P(t|x) := x ∈Nx |t∈T(x ) w(x, x ) x ∈Nx w(x, x ) where: Nx := {x ∈ X | (x, x ) ∈ R, T(x) = ∅} Runtime: O (|T||Nx |)) Leandro Balby Marinho 16 / 32 Machine Learning Lab, University of Hildesheim
  • 23. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Evaluation Datasets: dataset |U| |R| |T| Triples |Y | Posts |X| BibSonomy 116 361 412 10,148 2,522 Last.fm 2,917 1,853 2,045 219,702 75,565 Delicious 37,399 74,874 22,170 7,487,319 3,055,436 Evaluated methods: Baselines: (Locally) Constant Models (GCT,LCR, LCU). Ensemble of Locally Constant Models (LCE) [J¨aschke et al. 2008]. TopicRank, FolkRank [J¨aschke et al. 2007] RTF [Rendle et al. 2009] PITF [Rendle et al. 2010] Our NN-based Recommenders Leandro Balby Marinho 17 / 32 Machine Learning Lab, University of Hildesheim
  • 24. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Results: NN Methods 0 0.2 0.4 0.6 0.8 1 0 2 4 6 8 10 Recall Number of recommended tags Top-10 Tag Recommendations in Delicious WA CF UT CF UR matrixExt simEns LCR GCT Leandro Balby Marinho 18 / 32 Machine Learning Lab, University of Hildesheim
  • 25. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Results: WA vs. State-of-the-Art 0 0.2 0.4 0.6 0.8 1 0 2 4 6 8 10 Recall Number of recommended tags Top-10 Tag Recommendations in BibSonomy WA RTF PITF FolkRank LCE TopicRank Leandro Balby Marinho 19 / 32 Machine Learning Lab, University of Hildesheim
  • 26. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Results: WA vs. State-of-the-Art 0 0.2 0.4 0.6 0.8 1 0 2 4 6 8 10 Recall Number of recommended tags Top-10 Tag Recommendations in Last.fm PITF WA RTF FolkRank LCE TopicRank Leandro Balby Marinho 20 / 32 Machine Learning Lab, University of Hildesheim
  • 27. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Results: WA vs. State-of-the-Art 0 0.2 0.4 0.6 0.8 1 0 2 4 6 8 10 Recall Number of recommended tags Top-10 Tag Recommendations in Delicious PITF WA FolkRank LCE TopicRank Leandro Balby Marinho 21 / 32 Machine Learning Lab, University of Hildesheim
  • 28. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Runtime: WA vs. PITF BibSonomy Last.fm Delicious Method Runtime Runtime Runtime WA < 1 second < 1 minute ≈ 3 minutes PITF ≈ 5 minutes ≈ 7 hours ≈ 33 days Leandro Balby Marinho 22 / 32 Machine Learning Lab, University of Hildesheim
  • 29. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions ECML/Discovery Challenge 2009 2nd Place ECML/PKDD Discovery Challenge 2009! Rank Method Top-5 F1 1 PITF [Rendle et al. (2009)] 0.35594 2 Relational Ensemble [Marinho et al. (2009)]1 0.33185 – WA (not submitted) 0.32519 3 Content-based [Lipczak et al. (2009)] 0.32461 1 With Christine Preisach Leandro Balby Marinho 23 / 32 Machine Learning Lab, University of Hildesheim
  • 30. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Outline 1. Motivation 2. Problems and Contributions 3. Tag Recommender Systems 4. Nearest Neighbor-based Tag Recommendation 5. Cross-Tagging 6. Tag Enrichment 7. Conclusions and Future Work Leandro Balby Marinho 24 / 32 Machine Learning Lab, University of Hildesheim
  • 31. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Problem Use resources overlap to cross tags between systems. Leandro Balby Marinho 24 / 32 Machine Learning Lab, University of Hildesheim
  • 32. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Tag Recommendation for Cross-Tagging Cross-Tagging Approaches: LCR (locally constant per resource). Collaborative Filtering. Leandro Balby Marinho 25 / 32 Machine Learning Lab, University of Hildesheim
  • 33. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Evaluation Tag-Aware-based Evaluation The better the tags the better a tag-aware recommender that uses those tags. Tag-Aware based on HOSVD [Symeonidis et al. (2008)] Datasets Blogger.com Last.fm Annotated Blog |U| 6,620 44,143 3,827 |R| 17,372 17,372 1,323 |T| 0 4,903 422 |Y | 0 254,388 32,900 Leandro Balby Marinho 26 / 32 Machine Learning Lab, University of Hildesheim
  • 34. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Recall on the top-5 resources of HOSVD n - Number of tags used to annotate the test posts of Blogger.com. Leandro Balby Marinho 27 / 32 Machine Learning Lab, University of Hildesheim
  • 35. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Outline 1. Motivation 2. Problems and Contributions 3. Tag Recommender Systems 4. Nearest Neighbor-based Tag Recommendation 5. Cross-Tagging 6. Tag Enrichment 7. Conclusions and Future Work Leandro Balby Marinho 28 / 32 Machine Learning Lab, University of Hildesheim
  • 36. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Problems First we map tags from a folksonomy to concepts C of an ontology H : T → C Then we learn an ontology P such that: CP := T ˙∪ C The better the ontology the better a ontology-aware recommender that uses this ontology. Taxonomy driven CF [Ziegler et al. (2004)] Datasets: dataset |U| |T| |R| |Y | Last.fm 3,532 7,081 982 130,899 musicmoz - 555 982 - Leandro Balby Marinho 28 / 32 Machine Learning Lab, University of Hildesheim
  • 37. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Results 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Trivial Ontology Domain Expert Ontology Learned Ontology Recall Leandro Balby Marinho 29 / 32 Machine Learning Lab, University of Hildesheim
  • 38. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Outline 1. Motivation 2. Problems and Contributions 3. Tag Recommender Systems 4. Nearest Neighbor-based Tag Recommendation 5. Cross-Tagging 6. Tag Enrichment 7. Conclusions and Future Work Leandro Balby Marinho 30 / 32 Machine Learning Lab, University of Hildesheim
  • 39. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Conclusions Tag Sparsity: Nearest Neighbor Method that Performs competitively to more sophisticated methods. Require modest computational effort. Social Network Divide: Cross-tagging as a tag recommendation problem. Personalized cross-tagging better than non-personalized cross-tagging. Tag idiosyncrasy: Tag enrichment Well agreed concepts that match the semantic intention of users. Learned ontology better than trivial or domain expert ontology. New recommender systems-based evaluation protocols. Leandro Balby Marinho 30 / 32 Machine Learning Lab, University of Hildesheim
  • 40. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Future Work Optimzed weight learning for WA. Bidirectional Cross-Tagging. Optimized Cross-Tagging/Ontology learning. Leandro Balby Marinho 31 / 32 Machine Learning Lab, University of Hildesheim
  • 41. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Results NN vs. Baselines 0 0.2 0.4 0.6 0.8 1 0 2 4 6 8 10 Recall Number of recommended tags Top-10 Tag Recommendations in BibSonomy WA CF UT CF UR matrixExt simEns LCR GCT 0 0.2 0.4 0.6 0.8 1 0 2 4 6 8 10Recall Number of recommended tags Top-10 Tag Recommendations in Last.fm WA CF UT CF UR matrixExt simEns LCR GCT Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 42. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions PageRank for Folksonomias Based on PageRank [Hotho et al. 2006] Each hyperedge is broken into three undirected edges: Now PageRank can be applied: wt+1 ← λAT wt + (1 − λ)p Rank will be dominated by popular nodes (Skewd distribution of tag assignments) Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 43. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions FolkRank 1. First compute vector w(0) with p = 1. 2. Next compute vector w(1) with p[u] := 1 + |U|, p[r] := 1 + |R|, and p[v] := 1 for v = u, r. 3. Finally compute w := w(1) − w(0) . 4. Recommendation list ˆT(u, r) is the top-n nodes in the rank restricted to tags. Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 44. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions RTF: Ranking with Tensor Factorization Tag Recommendation as a tensor completion problem. Positive tags have higher rank than negative ones [Rendle et al. 2009]. yu,r,t1 > yu,r,t2 ⇔ (u, r, t1) ∈ T+ u,r ∧ (u, r, t2) ∈ T− u,r T+ u,r := {t | (u, r) ∈ Xtreino ∧ (u, r, t) ∈ Y }, T− u,r := {t | (u, r) ∈ Xtreino ∧ (u, r, t) Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 45. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Tucker Decomposition Model ˆY := ˆC ×u ˆU ×r ˆR ×t ˆT or equivalently: ˆyu,r,t = ˜u ˜r ˜t ˆc˜u,˜r,˜t · ˆuu,˜u · ˆrr,˜r · ˆtt,˜t where the model parameters are: ˆC ∈ RkU ×kR ×kT , ˆU ∈ R|U|×kU , ˆR ∈ R|R|×kR , ˆT ∈ R|T|×kT Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 46. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions PITF: Pairwise Interaction Tensor Factorization PITF only models the two-way interactions between user and tags as well as between resources and tags: ˆau,r,t = k f ˆuu,f · ˆtU t,f + k f ˆrr,f · ˆtR t,f where ˆU ∈ R|U|×k , ˆR ∈ R|R|×k , ˆTU ∈ R|T|×k and ˆTR ∈ R|T|×k Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 47. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Complexity Learning Runtime Complexity Method Runtime WA O(1) FolkRank O(1) RTF O iter · |Xtrain||T|2 · kU · kR · kT PITF O(iter · |Xtrain||T|2 · 2k) Prediction Runtime Complexity Method Runtime WA O (|T||Nx | + |T| log(n))) Folkrank O(iter · (|Y | + |U| + |R| + |T|) + |T| + |T| log(n)) RTF O(|T| · kU + kR · kT · kT ) PITF O(|T|2k + |T| log(n)) Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 48. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Relation Rewarding We can reward the best relation by a factor c ∈ R Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 49. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Results Cross-Tagging Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 50. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Tag Enrichment Approach Semantic mapping as an ontology matching problem. P(A, B) ≈ | A ∩ B | |R| [Doan et al. (2004)] Jaccard coefficient: JS(A, B) := P(A ∩ B)/P(A ∪ B) := P(A, B) P(A, B) + P(A, ¯B) + P(¯A, B) Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 51. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Ontology learning Frequent itemset mining for ontology learning [Marinho et al. 2008]2 . 2 Algorithm proposed by Krisztian Buza co-author of [Marinho et al. 2008] Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 52. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions Semantic mapping tags mapped concepts electro electronica hip hop hip hop chillout rock old skool dance house anything else but death heavy metal post-hardcore emo california punk political punk urban hip hop 60s stuff country-rock relaxing folk rock explorer experimental rock rock en espanol latin pop Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 53. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions An Extract of Domain Expert Ontology heavy_metal death_metal doom_metal black_metal thrash rap-metal hair_metal speed_metal grindcore metal Pajek Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim
  • 54. Motivation Problems Tag Recommender Systems NN Tag Recommendation Cross-Tagging Tag Enrichment Conclusions An Extract of Learned Ontology maynard james keenan powerful technical death metal brit-rock metalcore nu metal doom metal new age finnishprogressive metal alternative metal melodic death metal melancholic black metal progressive ethereal swedish metal gothenburg metal progressive death metal german bands i have seen live speed metal nwobhm heavy power metal symphonic metal guitargasm death-doom metal gothic metal famous frontman art rock viking metal groove metal melodic metal violent aggressive alternative - at work music moody faves a-o-t-w slipknot grindcore great lyrics gothenburg dark g00ds 70s progressive rock depressing cold doom art-rock prog trash metal depression brutal death metal us loud sad korn soad mezmerize fall out boy rap-metal seen them live nu-metal cello metal melodic black metal folk metal guitar music symphonic prog british metal awesome zeuhl female fronted metal love metal aggressive finland epic nellis1 symphonic black metal new metal ominous buen metal bands ive seen live classic thrash bands i have seen prog metal classic metal prog rock metal gods my band inspiration metal of some persuasionfavorite shitnice music grooving metal fav artistsblizzards main tags symphonic death grind melodic power metal everything speed favs melodic death heavy_metal death progressive_rock doom_metal death_metal thrash metal speed_metal periods Pajek Leandro Balby Marinho 32 / 32 Machine Learning Lab, University of Hildesheim