Upload
nicolas-marie
View
505
Download
0
Embed Size (px)
DESCRIPTION
Presentation at SEMANTICS2014: http://www.semantics.cc/www.semantics.cc/index.html Paper: Exploratory search on topics through different perspectives with DBpedia https://hal.inria.fr/hal-01057031 A promising scenario for combining linked data and search is exploratory search. During exploratory search, the search objective is ill-defined and favorable to discovery. A common limit of the existing linked data based exploratory search systems is that they constrain the exploration through single results selection and ranking schemes. The users can not influence the results to reveal specific aspects of knowledge that interest them. The models and algorithms we propose unveil such knowledge nuances by allowing the exploration of topics through several perspectives. The users adjust important computation parameters through three operations that help retrieving desired exploration perspectives: specification of interest criteria about the topic explored, controlled randomness injection to reveal unexpected knowledge and choice of the processed knowledge source(s). This paper describes the corresponding models, algorithms and the Discovery Hub implementation. It focuses on the three mentioned operations and presents their evaluations.
Citation preview
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Exploratory search on topics through differentperspectives with DBpedia
Nicolas Marie, Fabien Gandon, Alain Giboin, Γmilie Palagi
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
2
CONTEXTPROPOSITIONEVALUATIONCONCLUSION
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
3
CONTEXTPROPOSITIONEVALUATIONCONCLUSION
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Search is only a partially solved problem [White, 2009]Ambiguous queries, natural language queries, exploratory search tasksβ¦
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
10 blue links paradigm, simple, fast
Exploratory searchbottleneck
Discovery Hub
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Overviews
Faceted interfaces
Results clustering
Low-cost of browsing (going back-and-forth functionalities)
Query-suggestions and refinement
Serendipitous discoveries provocation
In-session of account related memory features
Exploratory search systems are optimized to support exploratory search tasks, common functionalities:
Discovery Hub
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery Hub
Linked data are promising for supporting exploratory search: β’ new algorithms β’ new interaction models optimized for exploration.
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery Hub
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
MaturityDiscovery Hub
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
1 topic of interest => 1 entity => 1 results set to explore
1 perspective
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery HubTopics are complex, multifaceted,
One entity => multiple perspectives &knowledge nuances
Entourage Art. movement
Curiositiesβ¦In French culture
In American culture
I want to discover Claude Monet (painter)...
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
MORE
Aemoo
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
13
CONTEXTPROPOSITIONEVALUATIONCONCLUSION
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery Hub
The models and algorithms we proposeunveil topic knowledge nuances by allowing the exploration of topics through several perspectives.
In the graph context of linked data these perspectives correspond to different non exclusive sets of objects and relations that are informative on a topic regarding specific aspects.
Flexible querying and data processing
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Refer to the papers for the complete formalization
Building perspectives thanks to spreading activation
β¦β¦
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery Hub
3 perspective-operations to expose knowledge nuances :
β’ Criteria of interest specificationβ’ Controlled randomness injectionβ’ Data source selection
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery Hub Criteria of interest specification
, dcterms:category, ?x
, dcterms:category, ?x
Classic similarity measure
, dcterms:category, ?a | ?b | ?c |...
, dcterms:category, ?a | ?b | ?c |...
Criteria spec. similarity
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery Hub Criteria of interest specification
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Classic β top 5 artists
Β« French / not impressonist Β» criteria specification β top 5 artists
Β« Not French / Impressonist Β» criteria specification β top 5 artists
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Chosen level of randomness
Randomness injection
* r + (1-r)*
* r + (1-r)*
* r + (1-r)*
* r + (1-r)*
* r + (1-r)*
* r + (1
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery Hub
Local Kgram instance
Data source selection
fr.dbpedia.org/sparql
it.dbpedia.org/sparql
de.dbpedia.org/sparql
es.dbpedia.org/sparql
dbpedia.org/sparql
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery HubData source selection
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
28
CONTEXTPROPOSITIONEVALUATIONCONCLUSION
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Evaluated algorithm versions
β’Basis algorithm of Discovery Hub
β’Personalized algorithm through criteria specification
β’Randomized algorithm, with 0.5 threshold
β’Highly randomized algorithm (Highly R.), with 1.0 threshold
Discovery Hub
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
β’ Hypothesis 1:
Users who specify their criteria of interest about a topic find the results of the search more relevant.
β’ Hypothesis 2:
Users who specify their criteria of interest about a topic do not find the results of the search less novel.
β’ Hypothesis 3:
The stronger is the level of randomness the more surprising the results are for the users.
β’ Hypothesis 4:
Even if the level of surprise is high, the majority of the top results are still relevant to the users.
Discovery Hub
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
π»3 π»ππβππ¦ π . ; ππ’ππππ πππ π ππ π’ππ‘ > π πππππππ§ππ ; ππ’πππππ πππ π ππ π’ππ‘ > [π΅ππ ππ ; ππ’πππππ πππ π ππ π’ππ‘]
[ " ; Suprising Relation] > " ; ππ’πππππ πππ π ππππ‘πππ > [ " ; ππ’πππππ πππ π ππππ‘πππ]
π»2 ππππ ππππππ§ππ ; ππ’ππππ πππ π ππ π’ππ‘ > [π΅ππ ππ ; ππ’πππππ πππ π ππ π’ππ‘]
[ " ; Surprising Relation] > [ " ; ππ’πππππ πππ π ππππ‘πππ]
π»1 ππππ ππππππ§ππ ; πΌππ‘ππππ π‘ > [π΅ππ ππ ; πΌππ‘ππππ π‘]
" ; π·ππ π‘ππππ < [ " ; π·ππ π‘ππππ]
π»4 π»ππβππ¦ π πππππππ§ππ ; πΌππ‘ππππ π‘ > π΄π£πππππ (2,5)
π πππππππ§ππ ; πΌππ‘ππππ π‘ > π΄π£πππππ (2,5)
(Highly R. : Highly Randomized)
Discovery Hub
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
A Β« good Β»
result in
ESS isβ¦
Our definitions Chosen metrics : Questions
(Likert Scale)
β¦ A
surprising
result
A result is surprising if :
β’ You discovered an
unknown resource or
relation
β’ You discovered
something unexpected
Surprising Result
This result is suprising
?
Surprising
Relation
This relation between
the topic searched and
the result is surprising
?
β¦ An
intersting
result
A result is interesting if :
β’ You think it is similar to
the topic explored
β’ You think you will
remind or reuse it
Interesting Result
This result is interesting
?
Distance between
the Result and the
topic searched
This result is too distant
from the topic searched
?
Discovery Hub
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
β’ 16 participants
β’ Phase 1 : Selection of 2 topics in a list of 20 queries randomly choose in the query log of Discovery Hub
- Information Visualization
- Serge Gainsbourg (french singer)
β’ Phase 2 : Specification of the categories of interest
β’ Phase 3 : User Test (~1h)
- Before the test
- Interview (name, age, do they know Discovery Hub ?,β¦)
- Presentation of Discovery Hub and the objective of the test
- Presentation of the questions and simulation
Discovery Hub
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
π»1 ππππ ππππππ§ππ ; πΌππ‘ππππ π‘ > [π΅ππ ππ ; πΌππ‘ππππ π‘]
" ; π·ππ π‘ππππ < [ " ; π·ππ π‘ππππ]
H1 : Users who specify their criteria of interest about a topic find the results of the search more relevant.
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
π»2 ππππ ππππππ§ππ ; ππ’ππππ πππ π ππ π’ππ‘ > [π΅ππ ππ ; ππ’πππππ πππ π ππ π’ππ‘]
[ " ; Surprising Relation] > [ " ; ππ’πππππ πππ π ππππ‘πππ]
H2: Users who specify their criteria of interest about a topic do not find the results of the search less novel
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
π»3 π»ππβππ¦ π . ; ππ’ππππ πππ π ππ π’ππ‘ > π πππππππ§ππ ; ππ’πππππ πππ π ππ π’ππ‘ > [π΅ππ ππ ; ππ’πππππ πππ π ππ π’ππ‘]
[ " ; Suprising Relation] > " ; ππ’πππππ πππ π ππππ‘πππ > [ " ; ππ’πππππ πππ π ππππ‘πππ]
(Highly R. : Highly Randomized)
H3: The stronger is the level of randomness the more surprising the results are for the users
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
π»4 π»ππβππ¦ π πππππππ§ππ ; πΌππ‘ππππ π‘ > π΄π£πππππ (2,5)
π πππππππ§ππ ; πΌππ‘ππππ π‘ > π΄π£πππππ (2,5)
H4: Even if the level of surprise is high, the majority of the top results are still relevant to the users
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
38
CONTEXTPROPOSITIONEVALUATIONCONCLUSION
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
Discovery Hub
β’ We proposed a framework to enable multi-perspective exploratory search:
- Formalization
- Implementation
- Evaluation
β’ 3 operators : criteria spec., randomization, data selection
β’ Evaluations globally positive, slight adjustements needed
β’ Interesting propositions from the reviewers, thank you
COPYRIGHT Β© 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.
http://semreco.inria.fr
Thank you ! Questions ?
Discovery Hub