----------------------------------------------
Apologies for cross-posting
----------------------------------------------
*******************************************************************************************
GenSW2017: 1st International Workshop on "Generalizing knowledge: from Machine Learning and
Knowledge Representation to the Semantic Web"
Website: https://sites.google.com/site/gensw2017/
In conjunction with with "The 16th International Conference of the Italian Association for
Artificial Intelligence" (AI*IA 2017), Bari, Italy, November 14 - 17 2017.
*******************************************************************************************
IMPORTANT DATES
-----------------------
Abstract submission: 13 July 2017
Paper submission: 18 July 2017
Notification to authors: 8 September 2017
Camera-ready copies: 29 September 2017
FOCUS
-----------------------
Generalizing descriptions is a problem traditionally investigated in at least two different fields
of Artificial Intelligence: Machine Learning (ML) and Knowledge Representation (KR). Both research
fields have played an important role in the development of the Semantic Web (SW).
KR provided the theoretical basis for formalizing shared knowledge bases, a.k.a. ontologies, and
for deductively reasoning over them. ML methods have been used for enriching ontologies, both at
schema and instance level, by exploiting inductive reasoning, while still benefiting from deductive
reasoning, when possible.
In the Web of Data, the availability of generalization mechanisms could be crucial for performing
several knowledge management tasks, such as data summarization, data indexing, cluster discovery and
many others. However, performing generalization in such a context cannot be done by just revisiting
traditional generalization services, because some issues and peculiarities need to be carefully
taken into account. One of these peculiarities is the data size, which requires new scalable
techniques. The second one is the data quality, which is affected by the endemic redundancy, noise,
frequent irrelevance and possible inconsistency of the available information. A third one is data
interdependency stemming from RDFS statements.
Despite some preliminary research efforts, very few solutions and methods can be found at the state
of the art for coping with this urgent problem. The maturity of solutions coming from the ML and KR
fields may certainly provide a reasonable starting point. However, methods mixing or stacking
solutions coming from both fields may result more promising to address all raised issues. Therefore,
the main goal of the workshop is to foster solutions cross-fertilizing both ML and KR fields,
focusing on generalizing SW knowledge descriptions and, possibly taking into account scalability
issues. Solutions of interest should cope with descriptions formalized, primarily, in RDF/RDFS, but
also in more expressive representation languages, like Description Logics/OWL.
The workshop aims at gathering solutions for the generalization of knowledge descriptions
formalized in standard representation languages for the Semantic Web (primarily, but not only,
RDF/RDFS). Solutions of interest should focus (primarily, but not only) on methods mixing and/or
stacking solutions coming from Machine Learning and Knowledge Representation fields and applicable
to standard Semantic Web representation languages. The developments of scalable solutions for this
purpose will be particularly appreciated.
TOPICS
--------------------------------------------------
Topics of interest include, but are not limited to:
· KR and/or ML methods (possibly in combination) for generalizing in the Semantic Web
· Semi-supervised, unbalanced, inductive learning for generalizing in the Semantic Web
· Reasoning services for generalization in the Semantic Web
- Generalization methods for finding commonalities and differences in the Semantic Web
- Generalization methods for enrichment Semantic Web knowledge bases
- Generalization methods for indexing in the Linked Data Cloud
· Evaluation and benchmarking of generalization approaches in the Semantic Web
· Scalable algorithms for generalizing the Web of Data
· Generalization in presence of uncertain/inconsistent/noisy knowledge
Papers should be written in English, formatted according to the Springer LNCS style, and not exceed
12 pages (full papers) or 4 pages (position papers) plus bibliography.
Papers must be submitted via easychair: https://easychair.org/conferences/?conf=gensw2017.
All accepted papers will be scheduled for oral presentations and will be published in CEUR Workshop
Proceedings AI*IA Series.
***************************************************************************************************
Authors of selected papers accepted to the workshop will be invited to submit an extended version
for publication on a SPECIAL ISSUE for the journal “Semantic Web – Interoperability, Usability,
Applicability" (http://www.semantic-web-journal.net/). Papers selected for the special issue have to
go through a full review process before acceptance.
***************************************************************************************************
ORGANIZING COMMITTEE
-----------------------------
Simona Colucci, Politecnico di Bari
Claudia d'Amato, Università degli Studi di Bari
Francesco M. Donini, Università della Tuscia, Viterbo
|