SlideShare una empresa de Scribd logo
1 de 38
Descargar para leer sin conexión
PERSONALIZED SEARCH FOR THE SOCIAL
SEMANTIC WEB
VLDB PHD WORKSHOP
Oana Tifrea-Marciuska
supervisor: Prof. Thomas Lukasiewicz
Department of Computer Science, University of Oxford, UK
August 31, 2015
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 1 /32
MOTIVATION
PRELIMINARIES
Datalog+/–
LANGUAGES
Strategies to Answer k-rank Disjunctive Atomic Queries
Experiments
FUTURE WORK
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 2 /32
WEB 3.0
Social Data Semantic data
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 3 /32
WEB 3.0 SEARCH
Social Data
Personalized access
Semantic data
Precise and rich results
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 3 /32
PERSONALIZED INFORMATION ACCESS
QUERY
ORDER BY user’s preferences
LIMIT k
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 4 /32
PERSONALIZED INFORMATION ACCESS
QUERY Datalog+/– Queries
ORDER BY user’s preferences Preference Model
LIMIT k Top k
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 4 /32
DATALOG+/–
hotel
id city conn class
t1 h1 rome c e
t2 h2 rome w l
t3 h3 rome c e
review
id user feedback
t7 h1 b n
t8 h2 b p
t9 h3 j p
reviewer
user age
t10 b 20
t11 j 30
friend
user user
t12 b a
t13 j a
FIGURE : Database D.
Constraints
Datalog like ∀X∀YΦ(X, Y) → Ψ(X)
friend(A, B) → friend(B, A)
With existential in the head ∀X∀YΦ(X, Y) → ∃ZΨ(X, Z)
reviewer(U, A) → ∃Ffriend(U, F)
Negative constraints ∀XΦ(X) → ⊥
friend(D, D) → ⊥
Equality constraints ∀XΦ(X) → Ai = Aj
review(A, U, F1) ∧ review(A, U, F2) → F1 = F2
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 5 /32
WHY DATALOG+/–
Generalizes DL-Lite family
Has implementations
Can use ideas from the preference modeling in database
community
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 6 /32
PERSONALIZED INFORMATION ACCESS
QUERY Datalog+/– Queries
ORDER BY user’s preferences Preference Model
LIMIT k Top k
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 7 /32
PREFERENCE MODELS
Syntax and Semantics
Advantages and Disadvantages: formal properties, experiments
Complexity for answering queries: conjunctive queries (CQ) or
disjunction of atomic queries(DAQ)
Solve conflict between data
Preferences of a group of users
Uncertainty (reviews) vs preferences
Desire: ontology language that handles preferences of user or a
group of users and can handle uncertainty (e.g., information
integration)
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 8 /32
CH 1. PP-DATALOG+/–∗
Combines (4 different operators)
Qualitative preferences - binary relations ⊆ HPref × HPref .
hotel(h1, rome, c, e) hotel(h2, rome, w, l)
Probabilistic model (e.g., reviews)
hotel(h3, rome, c, e) 0.9
hotel(h1, rome, c, e) 0.8
hotel(h2, rome, w, l) 0.3
Semantic properties
Complexity of answering top-k DAQ polynomial
Dataset: IMDB movie database, with synthetic preferences
∗In Journal on Data Semantics. Vol. 4. No. 2. Jun, 2015.
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 9 /32
CH. 1. GPP-DATALOG+/–∗
Combines for a group of people
Qualitative preferences - binary relations ⊆ HPref × HPref .
food(f1) food(f2)
Probabilistic model (e.g., reviews)
food(m2) 0.9; food(m3) 0.8; food(m2) 0.3
Semantic properties
Complexity of answering top-k DAQ polynomial
Dataset: YELP, with real preferences from 50 users
efficiency
quality of the results
∗In ACM Transactions on Internet Technology. Vol. 14. No. 4. Dec, 2014
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 10 /32
GROUP PREFERENCE MODEL
DEFINITION
A group preference model U = (U1, . . . , Un) for n 1 users is a collection
of n user preference models.
dest(f1
) dest(c3
) dest(c2
)
dest(c1
) dest(b1
) dest(f2
)
u1
dest(c1
)
dest(c3
)
u2
dest(f1
)
dest(b1
)dest(f2
)
dest(c2
)
dest(f1
)
u3
dest(b1
)
dest(c1
)
dest(f2
)dest(c3
)
dest(c2
)
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 11 /32
PROBABILISTIC MODEL
A preference relation is score-based if is defined as follows:
a1 a2 iff score(a1) > score(a2).
Model assigns a probability to each atom (using e.g. Markov
logic and Bayesian networks).
PrM
0.4
0.34
0.3
0.8
0.75
0.6
dest(b1
)
dest(c1
)
dest(f1
)
dest(c2
)
dest(f2
)
dest(c3
)
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 12 /32
CHALLENGES OF THE GIVEN MODEL 1/2
u1
dest(f1
) dest(c3
) dest(c2
)
dest(c1
) dest(b1
) dest(f2
)
u2
dest(f1
)
dest(c1
) dest(b1
)dest(f2
)
dest(c3
) dest(c2
)
dest(f1
)
u3
dest(b1
)
dest(c1
)
dest(f2
)dest(c3
)
dest(c2
)
0.8
0.75
0.6
dest(b1
)
dest(c1
)
dest(f1
)
PrM
0.4
0.34
0.3
dest(c2
)
dest(f2
)
dest(c3
)
Challenge 1: user preference model and the probabilistic model
in disagreement: preference merging operators
DEFINITION
Let U be an SPO and M be a score-based preference relation. A
preference merging operator ⊗( U, M ) yields a relation ∗
such that
1 ∗
is an SPO
2 if a1 U a2 and a1 M a2, then a1
∗
a2.
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 13 /32
CHALLENGES OF THE GIVEN MODEL 2/2
u1
dest(f1
) dest(c3
) dest(c2
)
dest(c1
) dest(b1
) dest(f2
)
u2
dest(f1
)
dest(c1
) dest(b1
)dest(f2
)
dest(c3
) dest(c2
)
dest(f1
)
u3
dest(b1
)
dest(c1
)
dest(f2
)dest(c3
)
dest(c2
)
0.8
0.75
0.6
dest(b1
)
dest(c1
)
dest(f1
)
PrM
0.4
0.34
0.3
dest(c2
)
dest(f2
)
dest(c3
)
Challenge 2: user preference models may be in disagreement
with each other: preference aggregation operator
DEFINITION
Let U = (U1, . . . , Un) be a group preference model, where every Ui is an
SPO. A preference aggregation operator on U yields an SPO ∗
.
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 14 /32
GPP-Datalog+/– ontology - our model
DEFINITION
A GPP-Datalog+/– ontology has the form KB = (O, U, M, ⊗, )
O is a Datalog+/– ontology
U = (U1, . . . , Un) is a group preference model with n 1
M is a probabilistic model
⊗ is a preference merging operator
is the preference aggregation operator
We say that KB is a guarded iff O is guarded.
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 15 /32
MERGING OPERATOR
0.4 − 0.6 > 0.1 No =⇒ keep relation
dest(f1
)
dest(c1
) dest(b1
)dest(f2
)
dest(c3
) dest(c2
)
0.8
0.75
0.6
dest(b1
)
dest(c1
)
dest(f1
)
PrM
0.4
0.34
0.3
dest(c2
)
dest(f2
)
dest(c3
)
dest(c1
) dest(b1
)
dest(f1
)
dest(c3
) dest(c2
) dest(f2
)
0.75 − 0.6 > 0.1 Yes =⇒ inverse relation
dest(f1
)
dest(c1
) dest(b1
)dest(f2
)
dest(c3
) dest(c2
)
0.8
0.75
0.6
dest(b1
)
dest(c1
)
dest(f1
)
PrM
0.4
0.34
0.3
dest(c2
)
dest(f2
)
dest(c3
)
dest(c1
) dest(b1
)
dest(f1
)
dest(c3
) dest(c2
) dest(f2
)
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 16 /32
SKYLINE AND K-RANK ANSWER
Let KB be a GPP-Datalog+/– ontology, Q(X) = q1(X1) ∨ · · · ∨ qn(Xn)
be a DAQ. Then, a skyline answer to Q relative to
∗
= (⊗( U1
, M ), . . . , ⊗( Un
, M )) is any θqi entailed by O such
that no θ exists with O |= θ qj and θ qj
∗
θqi , where θ and θ are
ground substitutions for the variables in Q(X).
A k-rank answer to Q is a sequence S = θ1, . . . θk built by
subsequently appending the skyline answers to Q, removing these
atoms from consideration, and repeating until either S = k or no more
skyline answers to Q remain.
dest(f1
)
dest(b1
)
dest(c1
)
dest(f2
)dest(c3
)
dest(c2
)
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 17 /32
STRATEGIES TO ANSWER K-RANK DAQ
Collapse to single user
1 Create virtual user
2 Calculate k-rank from it
Voting
1 Calculate k-rank for each of the users
2 Vote
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 18 /32
COLLAPSE TO SINGLE USER
no relation
dest(b1
)
u1
t = 0
dest(f2
)
dest(c3
)
dest(c1
) dest(c2
)
dest(f1
)
u2
t = 0.1
dest(c1
)
dest(f1
)
dest(c2
) dest(f2
)
dest(b1
)
dest(c3
)
u3
t = 0.19
dest(b1
)
dest(c1
)
dest(f1
)
dest(c3
)
dest(c2
)
dest(f2
) dest(b1
)
dest(c1
)
dest(f2
)
dest(c3
)
dest(c2
)
dest(f1
)
1
1
1
1
1
1
1
3
2
2
11 1
2
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 19 /32
COLLAPSE TO SINGLE USER
relation with weight 2
dest(b1
)
u1
t = 0
dest(f2
)
dest(c3
)
dest(c1
) dest(c2
)
dest(f1
)
u2
t = 0.1
dest(c1
)
dest(f1
)
dest(c2
) dest(f2
)
dest(b1
)
dest(c3
)
u3
t = 0.19
dest(b1
)
dest(c1
)
dest(f1
)
dest(c3
)
dest(c2
)
dest(f2
) dest(b1
)
dest(c1
)
dest(f2
)
dest(c3
)
dest(c2
)
dest(f1
)
1
1
1
1
1
1
1
3
2
2
11 1
2
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 20 /32
COLLAPSE TO SINGLE USER: K-RANK
Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 1
k-rank answer to Q
dest(b1) .
dest(c1
)
dest(f2
)
dest(f1
)
dest(c2
)
dest(b1
) 3
1
1
dest(c3
) 1
1
1
2
1
2
1
2
11
1
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
COLLAPSE TO SINGLE USER: K-RANK
Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 2
k-rank answer to Q
dest(b1) , dest(c1) .
dest(c1
)
dest(f2
)
dest(f1
)
dest(c2
)
1
dest(c3
) 1
1
2
1
2 11
1
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
COLLAPSE TO SINGLE USER: K-RANK
Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 3
k-rank answer to Q
dest(b1) , dest(c1) , dest(f1) .
dest(f2
)
dest(f1
)
dest(c2
)
1
dest(c3
) 1
11
1
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
COLLAPSE TO SINGLE USER: K-RANK
Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 4
k-rank answer to Q
dest(b1) , dest(c1) , dest(f1) , dest(f2) .
dest(f2
)dest(c2
)
dest(c3
) 1
1
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
COLLAPSE TO SINGLE USER: K-RANK
Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 5
k-rank answer to Q
dest(b1) , dest(c1) , dest(f1) , dest(f2) , dest(c2) or
dest(b1) , dest(c1) , dest(f1) , dest(f2) , dest(c3) .
dest(c2
)
dest(c3
)
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
VOTING - PLURALITY VOTING
Q = dest(X), k = 2, and (t1, t2, t3) = (0, 0.1, 0.19).
dest(f2) dest(c1) dest(c2) dest(b1) dest(c3)
u1 1 1 1 1 0
u2 0 1 0 1 0
u3 1 0 0 1 1
Total 2 2 1 3 1
dest(b1
)
u1
t = 0
dest(f2
)
dest(c3
)
dest(c1
) dest(c2
)
dest(f1
)
u3
t = 0.19
dest(b1
)
dest(c1
)
dest(f1
)
dest(c3
)
dest(c2
)
dest(f2
)
u2
t = 0.1
dest(c1
)
dest(f1
)
dest(c2
) dest(f2
)
dest(b1
)
dest(c3
)
k-rank answer to Q using plurality voting is dest(b1), dest(c1) or
dest(b1), dest(f2) .
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 22 /32
EXPERIMENTS
Yelp dataset: 1,000 businesses, 229,907 reviews to find places
to eat
categories in Yelp → Datalog+/–concepts (e.g Italian,
Mediterranean)
50 users inserted their preferences (e.g., the place to eat , food)
Compare methods
efficiency
quality of the results
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 23 /32
RESULTS OF THE EXPERIMENTS
group size increase → quality deacreases
k increase → quality increases
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 24 /32
CH 2. ONTOLOGICAL CP-THEORIES∗
A Datalog+/– ontology O and a set of υ: ξ ξ [W]
Given υ we prefer ξ to ξ , irrespective of the value of W, as long
as the other values are the same.
reviewer(bob, A): hotel(I, C, wifi, S) hotel(I, C, cable, S)[REVIEW]
with REVIEW = {review(p), review(n)}
Complexity of answering top-k CQ
∗In Proc. of the International Joint Conference on Artificial Intelligence, Jul 2015.
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 25 /32
ONTOLOGICAL CP-THEORIES
: review(I, U, p) review(I , U, n)[∅]
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 26 /32
ONTOLOGICAL CP-THEORIES
: hotel(I, C, w, S) hotel(I , C, c, S )[{R, F}]
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 27 /32
ONTOLOGICAL CP-THEORIES
hotel(I, C, c, S) review(I, U, n): reviewer(j, A) reviewer(beate, A )[∅]
:
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 28 /32
SUMMARY
Chapter Formalism ∗
1 PP-Datalog+/– [1]
1 GP-Datalog+/– [5]
1 GPP-Datalog+/–[2,3,4]
2 Ontological CP-nets [6,7]
2 Ontological CP-theories [8]
PP-Datalog+/-
GPP-Datalog+/-
Ontological CP-theories
Ontological CP-nets
1 2
GP-Datalog+/-
∗Complexity, Uncertainty, Properties and Implementation
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 29 /32
PUBLICATIONS
1 PP-Datalog+/– Preference–Based Query Answering in
Probabilistic Datalog+/– Ontologies In Journal on Data
Semantics. Vol. 4. No. 2. Pages 81–101. June, 2015.∗
2 Query Answering in Probabilistic Datalog+/– Ontologies under
Group Preferences. In Proc. of WI 2013. Pages 171–178.∗
3 Query Answering in Datalog+/– Ontologies under Group
Preferences and Probabilistic Uncertainty In Proc. DMSSW 2013
Vol. 8295 of LNCS. Pages 192–206.∗
4 Ontology–Based Query Answering with Group Preferences In
ACM Transactions on Internet Technology (TOIT). Vol. 14. No. 4.
Pages 25:1–25:24. December, 2014.∗
5 Group Preferences for Query Answering in Datalog+/–
Ontologies In Proc. of SUM 2013 Vol. 8078 of LNCS. Pages
360–373. 2013.∗
Journal, Conference, Workshop or symposium
∗Authors: T. Lukasiewicz, M.V. Martinez, G. I. Simari and O. Tifrea–Marciuska
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 30 /32
PUBLICATIONS
6 Computing k-Rank Answers with Ontological CP–Nets In Proc.
SEBD 2014. Pages 276-283.∗
7 Computing k-Rank Answers with Ontological CP–Nets In Proc.
PRUV 2014 Vol. 1205 of CEUR Workshop Proceedings. Pages
74–87.∗
8 Combining Existential Rules with the Power of CP–Theories In
Proc. of IJCAI 2015.∗
Conference, Workshop or symposium
∗Authors: T. Di Noia, T. Lukasiewicz, M.V. Martinez, G. I. Simari and O.
Tifrea–Marciuska
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 31 /32
THANK YOU
Questions ?oana.tifrea@cs.ox.ac.uk
OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 32 /32

Más contenido relacionado

Similar a VLDB Phd Workshop

Music Recommender Systems
Music Recommender SystemsMusic Recommender Systems
Music Recommender Systems
zhu02
 
breaking_dependencies_the_solid_principles__klaus_iglberger__cppcon_2020.pdf
breaking_dependencies_the_solid_principles__klaus_iglberger__cppcon_2020.pdfbreaking_dependencies_the_solid_principles__klaus_iglberger__cppcon_2020.pdf
breaking_dependencies_the_solid_principles__klaus_iglberger__cppcon_2020.pdf
VishalKumarJha10
 

Similar a VLDB Phd Workshop (20)

Chapter8-Link_Analysis.pptx
Chapter8-Link_Analysis.pptxChapter8-Link_Analysis.pptx
Chapter8-Link_Analysis.pptx
 
Chapter8-Link_Analysis (1).pptx
Chapter8-Link_Analysis (1).pptxChapter8-Link_Analysis (1).pptx
Chapter8-Link_Analysis (1).pptx
 
Music Recommender Systems
Music Recommender SystemsMusic Recommender Systems
Music Recommender Systems
 
breaking_dependencies_the_solid_principles__klaus_iglberger__cppcon_2020.pdf
breaking_dependencies_the_solid_principles__klaus_iglberger__cppcon_2020.pdfbreaking_dependencies_the_solid_principles__klaus_iglberger__cppcon_2020.pdf
breaking_dependencies_the_solid_principles__klaus_iglberger__cppcon_2020.pdf
 
Multichannel IoT CAUSAL digital twin
Multichannel IoT CAUSAL digital twinMultichannel IoT CAUSAL digital twin
Multichannel IoT CAUSAL digital twin
 
Graph Analytics with ArangoDB
Graph Analytics with ArangoDBGraph Analytics with ArangoDB
Graph Analytics with ArangoDB
 
Computer Scientists Retrieval - PDF Report
Computer Scientists Retrieval - PDF ReportComputer Scientists Retrieval - PDF Report
Computer Scientists Retrieval - PDF Report
 
Matrix Factorization
Matrix FactorizationMatrix Factorization
Matrix Factorization
 
Code dive 2019 kamil witecki - should i care about cpu cache
Code dive 2019   kamil witecki - should i care about cpu cacheCode dive 2019   kamil witecki - should i care about cpu cache
Code dive 2019 kamil witecki - should i care about cpu cache
 
Why functional programming and category theory strongly matters - Piotr Parad...
Why functional programming and category theory strongly matters - Piotr Parad...Why functional programming and category theory strongly matters - Piotr Parad...
Why functional programming and category theory strongly matters - Piotr Parad...
 
Why functional programming and category theory strongly matters
Why functional programming and category theory strongly mattersWhy functional programming and category theory strongly matters
Why functional programming and category theory strongly matters
 
What to do when one size does not fit all?!
What to do when one size does not fit all?!What to do when one size does not fit all?!
What to do when one size does not fit all?!
 
BlinkDB and G-OLA: Supporting Continuous Answers with Error Bars in SparkSQL-...
BlinkDB and G-OLA: Supporting Continuous Answers with Error Bars in SparkSQL-...BlinkDB and G-OLA: Supporting Continuous Answers with Error Bars in SparkSQL-...
BlinkDB and G-OLA: Supporting Continuous Answers with Error Bars in SparkSQL-...
 
Type and proof structures for concurrency
Type and proof structures for concurrencyType and proof structures for concurrency
Type and proof structures for concurrency
 
BlinkDB - Approximate Queries on Very Large Data
BlinkDB - Approximate Queries on Very Large DataBlinkDB - Approximate Queries on Very Large Data
BlinkDB - Approximate Queries on Very Large Data
 
Personalised Search for the Social Semantic Web
Personalised Search for the Social Semantic WebPersonalised Search for the Social Semantic Web
Personalised Search for the Social Semantic Web
 
A Generic Mapping-based Query Translation from SPARQL to Various Target Datab...
A Generic Mapping-based Query Translation from SPARQL to Various Target Datab...A Generic Mapping-based Query Translation from SPARQL to Various Target Datab...
A Generic Mapping-based Query Translation from SPARQL to Various Target Datab...
 
Reputation Systems I
Reputation Systems IReputation Systems I
Reputation Systems I
 
Overlapping community detection in Large-Scale Networks using BigCLAM model b...
Overlapping community detection in Large-Scale Networks using BigCLAM model b...Overlapping community detection in Large-Scale Networks using BigCLAM model b...
Overlapping community detection in Large-Scale Networks using BigCLAM model b...
 
Algebra
AlgebraAlgebra
Algebra
 

Más de Oana Tifrea-Marciuska

Google Anita Borg Memorial Scholarship - my story
Google Anita Borg Memorial Scholarship - my storyGoogle Anita Borg Memorial Scholarship - my story
Google Anita Borg Memorial Scholarship - my story
Oana Tifrea-Marciuska
 
Query Answering in Probabilistic Datalog+/{ Ontologies under Group Preferences
Query Answering in Probabilistic Datalog+/{ Ontologies under Group PreferencesQuery Answering in Probabilistic Datalog+/{ Ontologies under Group Preferences
Query Answering in Probabilistic Datalog+/{ Ontologies under Group Preferences
Oana Tifrea-Marciuska
 
The domain model of adaptive learning system - presentation
The domain model of adaptive learning system - presentationThe domain model of adaptive learning system - presentation
The domain model of adaptive learning system - presentation
Oana Tifrea-Marciuska
 

Más de Oana Tifrea-Marciuska (15)

Poster Presentation @ Oxford AI day
Poster Presentation @ Oxford AI dayPoster Presentation @ Oxford AI day
Poster Presentation @ Oxford AI day
 
Artificial intelligence for Social Good
Artificial intelligence for Social GoodArtificial intelligence for Social Good
Artificial intelligence for Social Good
 
Combining Existential Rules with the Power of CP-Theories
Combining Existential Rules with the Power of CP-TheoriesCombining Existential Rules with the Power of CP-Theories
Combining Existential Rules with the Power of CP-Theories
 
Combining Existential Rules with the Power of CP-Theories
Combining Existential Rules with the Power of CP-TheoriesCombining Existential Rules with the Power of CP-Theories
Combining Existential Rules with the Power of CP-Theories
 
OxBridge conference - my thesis in a minute
OxBridge conference - my thesis in a  minuteOxBridge conference - my thesis in a  minute
OxBridge conference - my thesis in a minute
 
Anita Borg birthday week 2015 @ Toplita - Romania
Anita Borg birthday week 2015 @ Toplita - RomaniaAnita Borg birthday week 2015 @ Toplita - Romania
Anita Borg birthday week 2015 @ Toplita - Romania
 
IMPROVING PERSONALIZED SEARCH ON SOCIAL WEB BASED ON SIMILARITIES BETWEEN USERS
IMPROVING PERSONALIZED SEARCH ON SOCIAL WEB BASED ON SIMILARITIES BETWEEN USERSIMPROVING PERSONALIZED SEARCH ON SOCIAL WEB BASED ON SIMILARITIES BETWEEN USERS
IMPROVING PERSONALIZED SEARCH ON SOCIAL WEB BASED ON SIMILARITIES BETWEEN USERS
 
Computing k-rank Answers with Ontological CP-nets
Computing k-rank Answers with Ontological CP-netsComputing k-rank Answers with Ontological CP-nets
Computing k-rank Answers with Ontological CP-nets
 
Group preferences in social networks
Group preferences in social networksGroup preferences in social networks
Group preferences in social networks
 
Query Answering in Probabilistic Datalog+/– Ontologies under Group Preferences
Query Answering in Probabilistic Datalog+/– Ontologies under Group PreferencesQuery Answering in Probabilistic Datalog+/– Ontologies under Group Preferences
Query Answering in Probabilistic Datalog+/– Ontologies under Group Preferences
 
Wi13 otm
Wi13 otmWi13 otm
Wi13 otm
 
Google Anita Borg Memorial Scholarship - my story
Google Anita Borg Memorial Scholarship - my storyGoogle Anita Borg Memorial Scholarship - my story
Google Anita Borg Memorial Scholarship - my story
 
Query Answering in Probabilistic Datalog+/{ Ontologies under Group Preferences
Query Answering in Probabilistic Datalog+/{ Ontologies under Group PreferencesQuery Answering in Probabilistic Datalog+/{ Ontologies under Group Preferences
Query Answering in Probabilistic Datalog+/{ Ontologies under Group Preferences
 
The domain model of adaptive learning system - presentation
The domain model of adaptive learning system - presentationThe domain model of adaptive learning system - presentation
The domain model of adaptive learning system - presentation
 
Fun Office
Fun OfficeFun Office
Fun Office
 

Último

Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak HamilCara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Kandungan 087776558899
 
Call Girls in Netaji Nagar, Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Netaji Nagar, Delhi 💯 Call Us 🔝9953056974 🔝 Escort ServiceCall Girls in Netaji Nagar, Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Netaji Nagar, Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
9953056974 Low Rate Call Girls In Saket, Delhi NCR
 
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort ServiceCall Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
9953056974 Low Rate Call Girls In Saket, Delhi NCR
 
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
ssuser89054b
 
VIP Call Girls Palanpur 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Palanpur 7001035870 Whatsapp Number, 24/07 BookingVIP Call Girls Palanpur 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Palanpur 7001035870 Whatsapp Number, 24/07 Booking
dharasingh5698
 
Call Now ≽ 9953056974 ≼🔝 Call Girls In New Ashok Nagar ≼🔝 Delhi door step de...
Call Now ≽ 9953056974 ≼🔝 Call Girls In New Ashok Nagar  ≼🔝 Delhi door step de...Call Now ≽ 9953056974 ≼🔝 Call Girls In New Ashok Nagar  ≼🔝 Delhi door step de...
Call Now ≽ 9953056974 ≼🔝 Call Girls In New Ashok Nagar ≼🔝 Delhi door step de...
9953056974 Low Rate Call Girls In Saket, Delhi NCR
 
Top Rated Call Girls In chittoor 📱 {7001035870} VIP Escorts chittoor
Top Rated Call Girls In chittoor 📱 {7001035870} VIP Escorts chittoorTop Rated Call Girls In chittoor 📱 {7001035870} VIP Escorts chittoor
Top Rated Call Girls In chittoor 📱 {7001035870} VIP Escorts chittoor
dharasingh5698
 

Último (20)

Employee leave management system project.
Employee leave management system project.Employee leave management system project.
Employee leave management system project.
 
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak HamilCara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
Cara Menggugurkan Sperma Yang Masuk Rahim Biyar Tidak Hamil
 
Call Girls in Netaji Nagar, Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Netaji Nagar, Delhi 💯 Call Us 🔝9953056974 🔝 Escort ServiceCall Girls in Netaji Nagar, Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Netaji Nagar, Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
 
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort ServiceCall Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
 
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
 
Water Industry Process Automation & Control Monthly - April 2024
Water Industry Process Automation & Control Monthly - April 2024Water Industry Process Automation & Control Monthly - April 2024
Water Industry Process Automation & Control Monthly - April 2024
 
A Study of Urban Area Plan for Pabna Municipality
A Study of Urban Area Plan for Pabna MunicipalityA Study of Urban Area Plan for Pabna Municipality
A Study of Urban Area Plan for Pabna Municipality
 
University management System project report..pdf
University management System project report..pdfUniversity management System project report..pdf
University management System project report..pdf
 
Hazard Identification (HAZID) vs. Hazard and Operability (HAZOP): A Comparati...
Hazard Identification (HAZID) vs. Hazard and Operability (HAZOP): A Comparati...Hazard Identification (HAZID) vs. Hazard and Operability (HAZOP): A Comparati...
Hazard Identification (HAZID) vs. Hazard and Operability (HAZOP): A Comparati...
 
Design For Accessibility: Getting it right from the start
Design For Accessibility: Getting it right from the startDesign For Accessibility: Getting it right from the start
Design For Accessibility: Getting it right from the start
 
Introduction to Serverless with AWS Lambda
Introduction to Serverless with AWS LambdaIntroduction to Serverless with AWS Lambda
Introduction to Serverless with AWS Lambda
 
KubeKraft presentation @CloudNativeHooghly
KubeKraft presentation @CloudNativeHooghlyKubeKraft presentation @CloudNativeHooghly
KubeKraft presentation @CloudNativeHooghly
 
DC MACHINE-Motoring and generation, Armature circuit equation
DC MACHINE-Motoring and generation, Armature circuit equationDC MACHINE-Motoring and generation, Armature circuit equation
DC MACHINE-Motoring and generation, Armature circuit equation
 
22-prompt engineering noted slide shown.pdf
22-prompt engineering noted slide shown.pdf22-prompt engineering noted slide shown.pdf
22-prompt engineering noted slide shown.pdf
 
VIP Call Girls Palanpur 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Palanpur 7001035870 Whatsapp Number, 24/07 BookingVIP Call Girls Palanpur 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Palanpur 7001035870 Whatsapp Number, 24/07 Booking
 
Navigating Complexity: The Role of Trusted Partners and VIAS3D in Dassault Sy...
Navigating Complexity: The Role of Trusted Partners and VIAS3D in Dassault Sy...Navigating Complexity: The Role of Trusted Partners and VIAS3D in Dassault Sy...
Navigating Complexity: The Role of Trusted Partners and VIAS3D in Dassault Sy...
 
Call Now ≽ 9953056974 ≼🔝 Call Girls In New Ashok Nagar ≼🔝 Delhi door step de...
Call Now ≽ 9953056974 ≼🔝 Call Girls In New Ashok Nagar  ≼🔝 Delhi door step de...Call Now ≽ 9953056974 ≼🔝 Call Girls In New Ashok Nagar  ≼🔝 Delhi door step de...
Call Now ≽ 9953056974 ≼🔝 Call Girls In New Ashok Nagar ≼🔝 Delhi door step de...
 
Thermal Engineering-R & A / C - unit - V
Thermal Engineering-R & A / C - unit - VThermal Engineering-R & A / C - unit - V
Thermal Engineering-R & A / C - unit - V
 
UNIT - IV - Air Compressors and its Performance
UNIT - IV - Air Compressors and its PerformanceUNIT - IV - Air Compressors and its Performance
UNIT - IV - Air Compressors and its Performance
 
Top Rated Call Girls In chittoor 📱 {7001035870} VIP Escorts chittoor
Top Rated Call Girls In chittoor 📱 {7001035870} VIP Escorts chittoorTop Rated Call Girls In chittoor 📱 {7001035870} VIP Escorts chittoor
Top Rated Call Girls In chittoor 📱 {7001035870} VIP Escorts chittoor
 

VLDB Phd Workshop

  • 1. PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB VLDB PHD WORKSHOP Oana Tifrea-Marciuska supervisor: Prof. Thomas Lukasiewicz Department of Computer Science, University of Oxford, UK August 31, 2015 OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 1 /32
  • 2. MOTIVATION PRELIMINARIES Datalog+/– LANGUAGES Strategies to Answer k-rank Disjunctive Atomic Queries Experiments FUTURE WORK OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 2 /32
  • 3. WEB 3.0 Social Data Semantic data OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 3 /32
  • 4. WEB 3.0 SEARCH Social Data Personalized access Semantic data Precise and rich results OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 3 /32
  • 5. PERSONALIZED INFORMATION ACCESS QUERY ORDER BY user’s preferences LIMIT k OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 4 /32
  • 6. PERSONALIZED INFORMATION ACCESS QUERY Datalog+/– Queries ORDER BY user’s preferences Preference Model LIMIT k Top k OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 4 /32
  • 7. DATALOG+/– hotel id city conn class t1 h1 rome c e t2 h2 rome w l t3 h3 rome c e review id user feedback t7 h1 b n t8 h2 b p t9 h3 j p reviewer user age t10 b 20 t11 j 30 friend user user t12 b a t13 j a FIGURE : Database D. Constraints Datalog like ∀X∀YΦ(X, Y) → Ψ(X) friend(A, B) → friend(B, A) With existential in the head ∀X∀YΦ(X, Y) → ∃ZΨ(X, Z) reviewer(U, A) → ∃Ffriend(U, F) Negative constraints ∀XΦ(X) → ⊥ friend(D, D) → ⊥ Equality constraints ∀XΦ(X) → Ai = Aj review(A, U, F1) ∧ review(A, U, F2) → F1 = F2 OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 5 /32
  • 8. WHY DATALOG+/– Generalizes DL-Lite family Has implementations Can use ideas from the preference modeling in database community OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 6 /32
  • 9. PERSONALIZED INFORMATION ACCESS QUERY Datalog+/– Queries ORDER BY user’s preferences Preference Model LIMIT k Top k OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 7 /32
  • 10. PREFERENCE MODELS Syntax and Semantics Advantages and Disadvantages: formal properties, experiments Complexity for answering queries: conjunctive queries (CQ) or disjunction of atomic queries(DAQ) Solve conflict between data Preferences of a group of users Uncertainty (reviews) vs preferences Desire: ontology language that handles preferences of user or a group of users and can handle uncertainty (e.g., information integration) OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 8 /32
  • 11. CH 1. PP-DATALOG+/–∗ Combines (4 different operators) Qualitative preferences - binary relations ⊆ HPref × HPref . hotel(h1, rome, c, e) hotel(h2, rome, w, l) Probabilistic model (e.g., reviews) hotel(h3, rome, c, e) 0.9 hotel(h1, rome, c, e) 0.8 hotel(h2, rome, w, l) 0.3 Semantic properties Complexity of answering top-k DAQ polynomial Dataset: IMDB movie database, with synthetic preferences ∗In Journal on Data Semantics. Vol. 4. No. 2. Jun, 2015. OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 9 /32
  • 12. CH. 1. GPP-DATALOG+/–∗ Combines for a group of people Qualitative preferences - binary relations ⊆ HPref × HPref . food(f1) food(f2) Probabilistic model (e.g., reviews) food(m2) 0.9; food(m3) 0.8; food(m2) 0.3 Semantic properties Complexity of answering top-k DAQ polynomial Dataset: YELP, with real preferences from 50 users efficiency quality of the results ∗In ACM Transactions on Internet Technology. Vol. 14. No. 4. Dec, 2014 OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 10 /32
  • 13. GROUP PREFERENCE MODEL DEFINITION A group preference model U = (U1, . . . , Un) for n 1 users is a collection of n user preference models. dest(f1 ) dest(c3 ) dest(c2 ) dest(c1 ) dest(b1 ) dest(f2 ) u1 dest(c1 ) dest(c3 ) u2 dest(f1 ) dest(b1 )dest(f2 ) dest(c2 ) dest(f1 ) u3 dest(b1 ) dest(c1 ) dest(f2 )dest(c3 ) dest(c2 ) OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 11 /32
  • 14. PROBABILISTIC MODEL A preference relation is score-based if is defined as follows: a1 a2 iff score(a1) > score(a2). Model assigns a probability to each atom (using e.g. Markov logic and Bayesian networks). PrM 0.4 0.34 0.3 0.8 0.75 0.6 dest(b1 ) dest(c1 ) dest(f1 ) dest(c2 ) dest(f2 ) dest(c3 ) OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 12 /32
  • 15. CHALLENGES OF THE GIVEN MODEL 1/2 u1 dest(f1 ) dest(c3 ) dest(c2 ) dest(c1 ) dest(b1 ) dest(f2 ) u2 dest(f1 ) dest(c1 ) dest(b1 )dest(f2 ) dest(c3 ) dest(c2 ) dest(f1 ) u3 dest(b1 ) dest(c1 ) dest(f2 )dest(c3 ) dest(c2 ) 0.8 0.75 0.6 dest(b1 ) dest(c1 ) dest(f1 ) PrM 0.4 0.34 0.3 dest(c2 ) dest(f2 ) dest(c3 ) Challenge 1: user preference model and the probabilistic model in disagreement: preference merging operators DEFINITION Let U be an SPO and M be a score-based preference relation. A preference merging operator ⊗( U, M ) yields a relation ∗ such that 1 ∗ is an SPO 2 if a1 U a2 and a1 M a2, then a1 ∗ a2. OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 13 /32
  • 16. CHALLENGES OF THE GIVEN MODEL 2/2 u1 dest(f1 ) dest(c3 ) dest(c2 ) dest(c1 ) dest(b1 ) dest(f2 ) u2 dest(f1 ) dest(c1 ) dest(b1 )dest(f2 ) dest(c3 ) dest(c2 ) dest(f1 ) u3 dest(b1 ) dest(c1 ) dest(f2 )dest(c3 ) dest(c2 ) 0.8 0.75 0.6 dest(b1 ) dest(c1 ) dest(f1 ) PrM 0.4 0.34 0.3 dest(c2 ) dest(f2 ) dest(c3 ) Challenge 2: user preference models may be in disagreement with each other: preference aggregation operator DEFINITION Let U = (U1, . . . , Un) be a group preference model, where every Ui is an SPO. A preference aggregation operator on U yields an SPO ∗ . OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 14 /32
  • 17. GPP-Datalog+/– ontology - our model DEFINITION A GPP-Datalog+/– ontology has the form KB = (O, U, M, ⊗, ) O is a Datalog+/– ontology U = (U1, . . . , Un) is a group preference model with n 1 M is a probabilistic model ⊗ is a preference merging operator is the preference aggregation operator We say that KB is a guarded iff O is guarded. OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 15 /32
  • 18. MERGING OPERATOR 0.4 − 0.6 > 0.1 No =⇒ keep relation dest(f1 ) dest(c1 ) dest(b1 )dest(f2 ) dest(c3 ) dest(c2 ) 0.8 0.75 0.6 dest(b1 ) dest(c1 ) dest(f1 ) PrM 0.4 0.34 0.3 dest(c2 ) dest(f2 ) dest(c3 ) dest(c1 ) dest(b1 ) dest(f1 ) dest(c3 ) dest(c2 ) dest(f2 ) 0.75 − 0.6 > 0.1 Yes =⇒ inverse relation dest(f1 ) dest(c1 ) dest(b1 )dest(f2 ) dest(c3 ) dest(c2 ) 0.8 0.75 0.6 dest(b1 ) dest(c1 ) dest(f1 ) PrM 0.4 0.34 0.3 dest(c2 ) dest(f2 ) dest(c3 ) dest(c1 ) dest(b1 ) dest(f1 ) dest(c3 ) dest(c2 ) dest(f2 ) OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 16 /32
  • 19. SKYLINE AND K-RANK ANSWER Let KB be a GPP-Datalog+/– ontology, Q(X) = q1(X1) ∨ · · · ∨ qn(Xn) be a DAQ. Then, a skyline answer to Q relative to ∗ = (⊗( U1 , M ), . . . , ⊗( Un , M )) is any θqi entailed by O such that no θ exists with O |= θ qj and θ qj ∗ θqi , where θ and θ are ground substitutions for the variables in Q(X). A k-rank answer to Q is a sequence S = θ1, . . . θk built by subsequently appending the skyline answers to Q, removing these atoms from consideration, and repeating until either S = k or no more skyline answers to Q remain. dest(f1 ) dest(b1 ) dest(c1 ) dest(f2 )dest(c3 ) dest(c2 ) OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 17 /32
  • 20. STRATEGIES TO ANSWER K-RANK DAQ Collapse to single user 1 Create virtual user 2 Calculate k-rank from it Voting 1 Calculate k-rank for each of the users 2 Vote OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 18 /32
  • 21. COLLAPSE TO SINGLE USER no relation dest(b1 ) u1 t = 0 dest(f2 ) dest(c3 ) dest(c1 ) dest(c2 ) dest(f1 ) u2 t = 0.1 dest(c1 ) dest(f1 ) dest(c2 ) dest(f2 ) dest(b1 ) dest(c3 ) u3 t = 0.19 dest(b1 ) dest(c1 ) dest(f1 ) dest(c3 ) dest(c2 ) dest(f2 ) dest(b1 ) dest(c1 ) dest(f2 ) dest(c3 ) dest(c2 ) dest(f1 ) 1 1 1 1 1 1 1 3 2 2 11 1 2 OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 19 /32
  • 22. COLLAPSE TO SINGLE USER relation with weight 2 dest(b1 ) u1 t = 0 dest(f2 ) dest(c3 ) dest(c1 ) dest(c2 ) dest(f1 ) u2 t = 0.1 dest(c1 ) dest(f1 ) dest(c2 ) dest(f2 ) dest(b1 ) dest(c3 ) u3 t = 0.19 dest(b1 ) dest(c1 ) dest(f1 ) dest(c3 ) dest(c2 ) dest(f2 ) dest(b1 ) dest(c1 ) dest(f2 ) dest(c3 ) dest(c2 ) dest(f1 ) 1 1 1 1 1 1 1 3 2 2 11 1 2 OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 20 /32
  • 23. COLLAPSE TO SINGLE USER: K-RANK Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 1 k-rank answer to Q dest(b1) . dest(c1 ) dest(f2 ) dest(f1 ) dest(c2 ) dest(b1 ) 3 1 1 dest(c3 ) 1 1 1 2 1 2 1 2 11 1 OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
  • 24. COLLAPSE TO SINGLE USER: K-RANK Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 2 k-rank answer to Q dest(b1) , dest(c1) . dest(c1 ) dest(f2 ) dest(f1 ) dest(c2 ) 1 dest(c3 ) 1 1 2 1 2 11 1 OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
  • 25. COLLAPSE TO SINGLE USER: K-RANK Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 3 k-rank answer to Q dest(b1) , dest(c1) , dest(f1) . dest(f2 ) dest(f1 ) dest(c2 ) 1 dest(c3 ) 1 11 1 OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
  • 26. COLLAPSE TO SINGLE USER: K-RANK Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 4 k-rank answer to Q dest(b1) , dest(c1) , dest(f1) , dest(f2) . dest(f2 )dest(c2 ) dest(c3 ) 1 1 OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
  • 27. COLLAPSE TO SINGLE USER: K-RANK Q = dest(X), (t1, t2, t3) = (0, 0.1, 0.19), k = 5 k-rank answer to Q dest(b1) , dest(c1) , dest(f1) , dest(f2) , dest(c2) or dest(b1) , dest(c1) , dest(f1) , dest(f2) , dest(c3) . dest(c2 ) dest(c3 ) OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 21 /32
  • 28. VOTING - PLURALITY VOTING Q = dest(X), k = 2, and (t1, t2, t3) = (0, 0.1, 0.19). dest(f2) dest(c1) dest(c2) dest(b1) dest(c3) u1 1 1 1 1 0 u2 0 1 0 1 0 u3 1 0 0 1 1 Total 2 2 1 3 1 dest(b1 ) u1 t = 0 dest(f2 ) dest(c3 ) dest(c1 ) dest(c2 ) dest(f1 ) u3 t = 0.19 dest(b1 ) dest(c1 ) dest(f1 ) dest(c3 ) dest(c2 ) dest(f2 ) u2 t = 0.1 dest(c1 ) dest(f1 ) dest(c2 ) dest(f2 ) dest(b1 ) dest(c3 ) k-rank answer to Q using plurality voting is dest(b1), dest(c1) or dest(b1), dest(f2) . OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 22 /32
  • 29. EXPERIMENTS Yelp dataset: 1,000 businesses, 229,907 reviews to find places to eat categories in Yelp → Datalog+/–concepts (e.g Italian, Mediterranean) 50 users inserted their preferences (e.g., the place to eat , food) Compare methods efficiency quality of the results OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 23 /32
  • 30. RESULTS OF THE EXPERIMENTS group size increase → quality deacreases k increase → quality increases OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 24 /32
  • 31. CH 2. ONTOLOGICAL CP-THEORIES∗ A Datalog+/– ontology O and a set of υ: ξ ξ [W] Given υ we prefer ξ to ξ , irrespective of the value of W, as long as the other values are the same. reviewer(bob, A): hotel(I, C, wifi, S) hotel(I, C, cable, S)[REVIEW] with REVIEW = {review(p), review(n)} Complexity of answering top-k CQ ∗In Proc. of the International Joint Conference on Artificial Intelligence, Jul 2015. OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 25 /32
  • 32. ONTOLOGICAL CP-THEORIES : review(I, U, p) review(I , U, n)[∅] OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 26 /32
  • 33. ONTOLOGICAL CP-THEORIES : hotel(I, C, w, S) hotel(I , C, c, S )[{R, F}] OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 27 /32
  • 34. ONTOLOGICAL CP-THEORIES hotel(I, C, c, S) review(I, U, n): reviewer(j, A) reviewer(beate, A )[∅] : OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 28 /32
  • 35. SUMMARY Chapter Formalism ∗ 1 PP-Datalog+/– [1] 1 GP-Datalog+/– [5] 1 GPP-Datalog+/–[2,3,4] 2 Ontological CP-nets [6,7] 2 Ontological CP-theories [8] PP-Datalog+/- GPP-Datalog+/- Ontological CP-theories Ontological CP-nets 1 2 GP-Datalog+/- ∗Complexity, Uncertainty, Properties and Implementation OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 29 /32
  • 36. PUBLICATIONS 1 PP-Datalog+/– Preference–Based Query Answering in Probabilistic Datalog+/– Ontologies In Journal on Data Semantics. Vol. 4. No. 2. Pages 81–101. June, 2015.∗ 2 Query Answering in Probabilistic Datalog+/– Ontologies under Group Preferences. In Proc. of WI 2013. Pages 171–178.∗ 3 Query Answering in Datalog+/– Ontologies under Group Preferences and Probabilistic Uncertainty In Proc. DMSSW 2013 Vol. 8295 of LNCS. Pages 192–206.∗ 4 Ontology–Based Query Answering with Group Preferences In ACM Transactions on Internet Technology (TOIT). Vol. 14. No. 4. Pages 25:1–25:24. December, 2014.∗ 5 Group Preferences for Query Answering in Datalog+/– Ontologies In Proc. of SUM 2013 Vol. 8078 of LNCS. Pages 360–373. 2013.∗ Journal, Conference, Workshop or symposium ∗Authors: T. Lukasiewicz, M.V. Martinez, G. I. Simari and O. Tifrea–Marciuska OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 30 /32
  • 37. PUBLICATIONS 6 Computing k-Rank Answers with Ontological CP–Nets In Proc. SEBD 2014. Pages 276-283.∗ 7 Computing k-Rank Answers with Ontological CP–Nets In Proc. PRUV 2014 Vol. 1205 of CEUR Workshop Proceedings. Pages 74–87.∗ 8 Combining Existential Rules with the Power of CP–Theories In Proc. of IJCAI 2015.∗ Conference, Workshop or symposium ∗Authors: T. Di Noia, T. Lukasiewicz, M.V. Martinez, G. I. Simari and O. Tifrea–Marciuska OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 31 /32
  • 38. THANK YOU Questions ?oana.tifrea@cs.ox.ac.uk OANA TIFREA-MARCIUSKA PERSONALIZED SEARCH FOR THE SOCIAL SEMANTIC WEB SLIDE 32 /32