A Generic Semantic-based Framework
for Cross-domain Recommendation
Ignacio Fernández-Tobías1, Marius Kaminskas2, Iván Cantador1, Francesco Ricci2 1 Escuela Politécnica Superior, Universidad Autónoma de Madrid, Spain [email protected], [email protected]
2 Faculty of Computer Science, Free University of Bozen-Bolzano, Italy [email protected], [email protected]
1
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
• Cross-domain recommendation
• Case study: adapting music recommendation to points of interest
• A semantic-based framework for cross-domain recommendation
• Semantic-based knowledge representation
• Semantic graph-based recommendation algorithm
• Preliminary results
• Future work
Contents
2
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
• Cross-domain recommendation
• Case study: adapting music recommendation to points of interest
• A semantic-based framework for cross-domain recommendation
• Semantic-based knowledge representation
• Semantic graph-based recommendation algorithm
• Preliminary results
• Future work
Contents
3
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Cross-domain recommendation
• Recommender systems can help users to make choices, by proactively
finding relevant items or services, taking into account or predicting the
users’ tastes, priorities and goals
• The vast majority of the currently available recommender systems predict
the user’s relevance of items in a specific and limited domain
4
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Cross-domain recommendation
• In some applications, it could be useful to offer the user joint personalized
recommendations of items belonging to multiple domains
• In an e-commerce site, we may suggest movies or videogames based on a particular
book bought by a costumer
• In a travel application, we may suggest cultural events may interest a person who has
booked a hotel in a particular place
• In an e-learning system, we may suggest educational websites with topics related to a
video documentary a student has seen
• Potential benefits
• Offering diversity and serendipity
• Addressing the user cold-start problem (on the target domain)
• Mitigating the sparsity problem
5
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Cross-domain recommendation
• Some real applications do already recommend items from different
domains, but
• their recommendations rely on statistical analysis of popular items, without any
personalization strategy, or
• most of them only exploit information about the user preferences available in the target
domain
6
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Cross-domain recommendation
• Research questions [Winoto & Tang, 2008]
1. At community level, are there correlations between user preferences for items
belonging to the different domains of interest?
2. At individual level, can we build a recommendation model where each user’s
preferences in source domains are used to predict/adapt her preferences in target
domains?
3. How should we evaluate the effectiveness of cross-domain item recommendations?
[Winoto & Tang, 2008] Winoto, P., Tang, T. 2008. If You Like the Devil Wears Prada the Book, Will You also Enjoy the Devil Wears
Prada the Movie? A Study of Cross-Domain Recommendations. New Generation Computing 26(3), 209-225.
7
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
• Cross-domain recommendation
• Case study: adapting music recommendation to points of interest
• A semantic-based framework for cross-domain recommendation
• Semantic-based knowledge representation
• Semantic graph-based recommendation algorithm
• Preliminary results
• Future work
Contents
8
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
• Recommending music artists that suit places of interest (POIs)
• Mobile city guide soundtrack
• Adaptive music playlist in a car
Case study: adapting music recommendation to points of interest
[Braunhofer et al., 2011] Braunhofer, M., Kaminskas, M., Ricci, F. 2011. Recommending Music for Places of Interest in a Mobile
Travel Guide. 5th ACM Conference on Recommender Systems.
9
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Case study: adapting music recommendation to points of interest
• In a previous work [Kaminskas & Ricci, 2011], emotional tags were used to
manually annotate places and music
• Emotional tags can be used to find matching between music and places of interest
‐ e.g. a monument and a music track may be described as ‘strong’ and ‘triumphant’
[Kaminskas & Ricci, 2011] Kaminskas, M., Ricci, F. 2011. Location-Adapted Music Recommendation Using Tags. 19th International
Conference on User Modeling, Adaptation and Personalization, 183-194.
10
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Case study: adapting music recommendation to points of interest
• In this work, we aim at automatically finding semantic relations between
POIs and music artists
• We propose to explore the Web of Data (Linked
Data) to find such relations
• Specifically, we propose to exploit DBpedia, the
Linked Data version of Wikipedia
• DBpedia can be considered as a core ontology in
the Web of Data
• Connected to many other ontologies
• Describing and linking more than 3.5 million
concepts from a large variety of knowledge
domains
11
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Case study: adapting music recommendation to points of interest
• In this work, we aim at automatically finding semantic relations between
POIs and music artists
• We propose to explore the Web of Data (Linked
Data) to find such relations
• Specifically, we propose to exploit DBpedia, the
Linked Data version of Wikipedia
• DBpedia can be considered as a core ontology in
the Web of Data
• Connected to many other ontologies
• Describing and linking more than 3.5 million
concepts from a large variety of knowledge
domains
12
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Case study: adapting music recommendation to points of interest
• Issues to investigate, identified in [Winoto & Tang, 2008]
1. Correlations between user preferences for items of the different domains
Correlations between POIs and music were established through tags in
[Kaminskas & Ricci, 2011]
2. Recommendation model to predict/adapt user preferences across domains
This paper addresses this particular issue, presenting a semantic-based
framework to support cross-domain recommendation
3. Evaluation of cross-domain recommendation effectiveness
Future work
13
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
• Cross-domain recommendation
• Case study: adapting music recommendation to points of interest
• A semantic-based framework for cross-domain recommendation
• Semantic-based knowledge representation
• Semantic graph-based recommendation algorithm
• Preliminary results
• Future work
Contents
14
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
• Goal: finding semantic relations between a given POI and music artists
• Example: music artists related to the ‘Vienna State Opera’
• Identified relations:
• Geographical: artists who were born, died or lived in Vienna
• Time-based: artists who were born, died or lived in the year (decade, century) the
State Opera of Vienna was built
• Category-based: artists who belong to music categories that are related through
keywords to architecture structures/styles identified with the building of the Opera of
Vienna
• Tags: artists annotated with tags also assigned to the Opera of Vienna
A Semantic-based framework for cross-domain recommendation
Vienna State Opera Wolfgang Amadeus Mozart
15
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
A Semantic-based framework for cross-domain recommendation
• A directed Acyclic Graph (DAG) representing semantic relations between
concepts in two domains
State Opera
of Vienna
State Opera
of Vienna
Vienna Austria Vienna Austria
19th century
19th century
Opera houses Opera houses
opera opera Opera
composers Opera
composers
Mozart Mozart
Brahms Brahms Bizet Bizet
Ballet venues Ballet
venues ballet ballet
Ballet composers
Ballet composers
Arnold Schoenberg
Arnold Schoenberg
POI
CITY
TIME
ARCHITECTURE CATEGORY
KEYWORD MUSIC CATEGORY
MUSIC ARTIST instance
class
16
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
A Semantic-based framework for cross-domain recommendation
• The previous graph can be considered as a particular instance of a
semantic class/category network
• The selection of classes and relations is guided by experts on the domains
of interest and knowledge repositories
POI POI
CITY CITY
TIME TIME
ARCHITECTURE CATEGORY
ARCHITECTURE CATEGORY
KEYWORD KEYWORD MUSIC
CATEGORY MUSIC
CATEGORY
MUSIC ARTIST MUSIC ARTIST
located in
was built
belongs to
subcategory of subcategory of
was born, died, lived in
was born, died, lived in
has keyword keyword of
17
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
A Semantic-based framework for cross-domain recommendation
• As a proof of concept, we have built our approach by exploiting DBpedia
ontology in two stages:
1. Manually identifying DBpedia classes and relations belonging to the domains of
interest to define the semantic-based knowledge representation
2. Automatically obtaining related DBpedia instances according to the classes and
relations identified in the first stage
POI POI MUSIC ARTIST MUSIC ARTIST
Semantic framework Semantic network 1 2
Vienna State Opera
Wolfgang Amadeus Mozart
18
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
• Cross-domain recommendation
• Case study: adapting music recommendation to points of interest
• A semantic-based framework for cross-domain recommendation
• Semantic-based knowledge representation
• Semantic graph-based recommendation algorithm
• Preliminary results
• Future work
Contents
19
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Semantic graph-based recommendation algorithm
• In the semantic network, a final score for each concept can be computed by
weight spreading strategies
• Initial weight values for concepts and relations must be established
State Opera
of Vienna
State Opera
of Vienna
Vienna Austria Vienna Austria
19th century
19th century
Opera houses Opera houses
opera opera Opera
composers Opera
composers
Mozart Mozart
Brahms Brahms Bizet Bizet
Ballet venues Ballet
venues ballet ballet
Ballet composers
Ballet composers
Arnold Schoenberg
Arnold Schoenberg
1
1
0.3
0.5
0.5
1
1
0.3
0.4
0.4
0.4
0.4
0.6
0.6
0.6
0.3
0.6
20
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Semantic graph-based recommendation algorithm
State Opera
of Vienna
State Opera
of Vienna
Vienna Austria Vienna Austria
19th century
19th century
Opera houses Opera houses
opera opera Opera
composers Opera
composers
Mozart Mozart
Brahms Brahms
Bizet Bizet
Ballet venues Ballet
venues ballet ballet
Ballet composers
Ballet composers
Arnold Schoenberg
Arnold Schoenberg
1
1
0.3
0.5
0.5
1
1
0.3
0.4
0.4
0.4
0.4
0.6
0.6 0.6
0.6
1·1=1
1·0.3=0.3
1·0.5=0.5
1·0.5=0.5
0.3
21
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Semantic graph-based recommendation algorithm
State Opera
of Vienna
State Opera
of Vienna
Vienna Austria Vienna Austria
19th century
19th century
Opera houses Opera houses
opera opera Opera
composers Opera
composers
Mozart Mozart
Brahms Brahms
Bizet Bizet
Ballet venues Ballet
venues ballet ballet
Ballet composers
Ballet composers
Arnold Schoenberg
Arnold Schoenberg
1
1
0.3
0.5
0.5
1
1
0.3
0.4
0.4
0.4
0.4
0.6
0.6 0.6
0.3
0.5
0.5
0.5·0.4=0.2
0.5·0.4=0.2
0.3
1
0.6
22
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Semantic graph-based recommendation algorithm
0.2·0.4=0.08
State Opera
of Vienna
State Opera
of Vienna
Vienna Austria Vienna Austria
19th century
19th century
Opera houses Opera houses
opera opera Opera
composers Opera
composers
Mozart Mozart
Brahms Brahms
Bizet Bizet
Ballet venues Ballet
venues ballet ballet
Ballet composers
Ballet composers
Arnold Schoenberg
Arnold Schoenberg
1
1
0.3
0.5
0.5
1
1
0.3
0.4
0.4
0.4
0.4
0.6
0.6 0.6
0.6
0.2
0.2 0.2·0.4=0.08
0.3
0.3
0.5
0.5
1
23
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Semantic graph-based recommendation algorithm
0.08
State Opera
of Vienna
State Opera
of Vienna
Vienna Austria Vienna Austria
19th century
19th century
Opera houses Opera houses
opera opera Opera
composers Opera
composers
Mozart Mozart
Brahms Brahms
Bizet Bizet
Ballet venues Ballet
venues ballet ballet
Ballet composers
Ballet composers
Arnold Schoenberg
Arnold Schoenberg
1
1
0.3
0.5
0.5
1
1
0.3
0.4
0.4
0.4
0.4
0.6
0.6 0.6
0.6
0.08
1·1+0.08·0.6+0.08·0.6+0.3·0.3=1.186
0.08·0.6=0.048
1·1+0.08·0.6=1.048
0.3·0.3=0.09
0.3
0.3
0.5
0.5
1
0.2
0.2
24
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Semantic graph-based recommendation algorithm
0.08
State Opera
of Vienna
State Opera
of Vienna
Vienna Austria Vienna Austria
19th century
19th century
Opera houses Opera houses
opera opera Opera
composers Opera
composers
Mozart Mozart
Brahms Brahms
Bizet Bizet
Ballet venues Ballet
venues ballet ballet
Ballet composers
Ballet composers
Arnold Schoenberg
Arnold Schoenberg
1
1
0.3
0.5
0.5
1
1
0.3
0.4
0.4
0.4
0.4
0.6
0.6 0.6
0.6
0.08·0.6=0.048
1·1+0.08·0.6=1.048
0.3·0.3=0.09
1·1+0.08·0.6+0.08·0.6+0.3·0.3=1.186
0.3
0.3
0.5
0.5
1
0.2
0.2 0.08
25
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Semantic graph-based recommendation algorithm
• The initial weights of an edge in the graph can depend on the relevance of
the linked instances and of the corresponding semantic classes
• These relevance values could be assigned in different ways
),(rel),',(rel)',( 'r IIr CCIIfIIV
Class relevance Domain expert
e.g. a city is more informative to link a POI than a keyword
Instance relevance User profile
e.g. an interest in Mozart’s compositions the relevance
for Mozart gets higher
Relation relevance Entity semantic similarity
e.g. co-occurrences of concepts ‘Mozart’ and ‘Vienna’ within a
document collection
26
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Semantic graph-based recommendation algorithm
• In general, the weight of an instance not only depends on its relevance
value and that of its class, but also inductively on the weights of the
predecessors in the network
kII ,,1
),(,),,();(,),();(rel),(rel)( 11ee IIVIIVIWIWCIgIW kkI
]1,0[),,(rel)1()',(rel)',( 'rr II CCIIIIV
k
p
Ipp CIIVIWIW1
e ]1,0[),(rel)1(),()()(
• To preliminarily test our approach we have implemented a simple retrieval
algorithm computing weights by linear combination
27
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
• Cross-domain recommendation
• Case study: adapting music recommendation to points of interest
• A semantic-based framework for cross-domain recommendation
• Semantic-based knowledge representation
• Semantic graph-based recommendation algorithm
• Preliminary results
• Future work
Contents
28
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Preliminary results
• Example: ‘Vienna State Opera’ (Vienna, Austria)
29
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Preliminary results
• Top 10 musicians for ‘Vienna State Opera’
Music artist Top music genres Born/Death countries Date
Arnold Schoenberg Classical
Avant-garde
Austria
USA 20th century
Wolfgang Amadeus Mozart Classical
Instrumental
Austria
Austria 18th century
Emil von Reznicek Classical
Opera
Austria
Germany 20th century
Alban Berg Classical
Contemporary
Hungary
Austria 20th century
Ludwig van Beethoven Classical
Instrumental
Germany
Austria 19th century
Antonio Vivaldi Classical
Baroque
Italy
Austria 18th century
Giovanni Felice Sances Classical
Baroque
Italy
Austria 17th century
Fritz Kreisler Classical
Violin
Austria
USA 20th century
Georg Christoph Wagenseil Classical
Baroque
Austria
Austria 18th century
Antonio Salieri Classical
Italian
Italy
Austria 19th century
30
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Preliminary results
• Example: found relations between ‘Vienna State Opera’ and ‘Wolfgang
Amadeus Mozart’ PLACE OF INTEREST: Vienna State Opera
CITY: Vienna, Austria
MUSIC ARTIST: Wolfgang Amadeus Mozart
ARCHITECTURE CATEGORY: Opera houses
KEYWORD: opera
MUSIC CATEGORY: Opera composers
MUSIC ARTIST: Wolfgang Amadeus Mozart
TAG: energetic
MUSIC CATEGORY: Opera composers
MUSIC ARTIST: Wolfgang Amadeus Mozart
TAG: sentimental
MUSIC CATEGORY: Opera composers
MUSIC ARTIST: Wolfgang Amadeus Mozart
MUSIC GENRE: classical
MUSIC ARTIST: Wolfgang Amadeus Mozart
ARCHITECTURE CATEGORY: Theatres
TAG: animated
MUSIC GENRE: classical
MUSIC ARTIST: Wolfgang Amadeus Mozart
31
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Preliminary results
• Example: ‘Wembley Stadium’ (London, UK)
32
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Preliminary results
• Top 10 musicians for ‘Wembley Stadium’
Music artist Top music genres Born/Death Countries Date
Beady Eye
(Oasis band members)
Rock
British
UK
(origin) 2009
Operahouse Indie Rock
British
UK
(origin) 2006
The Woe Betides Rock
Grunge
UK
(origin) 2008
Skunk Anansie Rock
Female vocalist
UK
(origin) 1994
The Fallen Leaves Garage
Acoustic
UK
(origin) 2004
Ivyrise Rock
Alternative
UK
(origin) 2007
Plastic Ono Band
(John Lennon & Yoko Ono)
Experimental
Avant-garde
UK
(origin) 1969
We Are Balboa Indie Rock
Female vocalist
Spain-UK
(origin) 2003
Goldhawks Rock
British
UK
(origin) 2009
Teddy Thompson Folk
British
UK
USA 1976
33
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Preliminary results
PLACE OF INTEREST: Wembley Stadium
CITY: London, United Kingdom
MUSIC ARTIST: Beady Eye
TIME: 2007
MUSIC ARTIST: Beady Eye
ARCHITECTURE CATEGORY: Music venues
ARCHITECTURE CATEGORY: Rock music venues
KEYWORD: rock
MUSIC CATEGORY: Indie rock
MUSIC ARTIST: Beady Eye
MUSIC CATEGORY: Rock music
MUSIC ARTIST: Beady Eye
TAG: strong
MUSIC CATEGORY: Rock music
MUSIC ARTIST: Beady Eye
• Example: found relations between ‘Wembley Stadium’ and ‘Beady Eye’
34
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Preliminary results
• Automatic extraction of data from DBPedia for an input city
• Modular and extensible implementation of the framework
• Dataset
• 3098 POIs located in 21 European cities
‐ 147.5 POIs/city
• 697 architecture categories
‐ 229 are directly linked to POIs
‐ Avg. 1.4 categories/POI
• 109 keywords describing 181 different architecture categories
‐ Avg. 1.1 keywords/category
• 1568 music artists
• 1116 music categories
‐ 309 directly linked to artists (avg. 1.7 categories/artist)
‐ 511 related to keywords (avg. 1.2 keywords/category)
• Time data for 64.72% of the POIs
35
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
• Cross-domain recommendation
• Case study: adapting music recommendation to points of interest
• A semantic-based framework for cross-domain recommendation
• Semantic-based knowledge representation
• Semantic graph-based recommendation algorithm
• Preliminary results
• Future work
Contents
36
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Future work
• Evaluation – user study
• Are semantically relations between POIs and music artists really appreciated by users
in a recommendation scenario?
• Do users find cross-domain recommendations meaningful, and prefer them over non-
adapted music suggestions?
• Providing personalized recommendations
• Cascade strategy
‐ Obtaining semantically related artists to the input POI
‐ Ranking (adding, removing) artists with a recommender based on the user’s
preferences
37
A Generic Semantic-based Framework for Cross-domain Recommendation
2nd Workshop on Information Heterogeneity and Fusion in Recommender Systems (HetRec 2011)
5th ACM Conference on Recommender Systems (RecSys 2011) - Chicago, IL, USA - October 23-27, 2011
Future work
• Initializing entity and relation weights
• Exploiting data statistics to estimate the popularity of the semantic entities and
relations
• Exploring several weight spreading strategies
• Constrained Spreading Activation
‐ Node in/out degrees
‐ Weight propagation thresholds
‐ Path length thresholds
• Flow Networks
‐ Ford-Fulkerson’s algorithm to find maximum network flow
• Semi-automatic defining the semantic framework
• Automatically exploring DBpedia to identify relevant entities and relations describing
the domains of interest
A Generic Semantic-based Framework
for Cross-domain Recommendation
Ignacio Fernández-Tobías1, Marius Kaminskas2, Iván Cantador1, Francesco Ricci2 1 Escuela Politécnica Superior, Universidad Autónoma de Madrid, Spain [email protected], [email protected]
2 Faculty of Computer Science, Free University of Bozen-Bolzano, Italy [email protected], [email protected]