W3C home > Mailing lists > Public > semantic-web@w3.org > August 2014

CfP: IJSWIS Special Issue on Dataset Profiling and Federated Search for Linked Data

From: Stefan Dietze <dietze@l3s.de>
Date: Tue, 26 Aug 2014 12:35:36 +0200
Message-ID: <53FC62F8.5030204@l3s.de>
To: Public LOD community <public-lod@w3.org>, "semantic-web@w3.org" <semantic-web@w3.org>
[Apologies for cross-posting]


=> Special issue website at http://www.ijswis.org/?q=node/51
=> Submission by 8 December 2014


While the Web of Data, and in particular Linked Data, has seen 
tremendous growth over the past years, take-up, usage and reuse of data 
is still limited and is often focused on well-known reference datasets. 
The main obstacles preventing users from obtaining relevant, correct and 
up-to-date information from distributed LOD datasets is the lack of 
scalable and usable methods for formulating and distributing semantic 
and keyword queries across the Web of Data. This problem is  further 
alleviated by the lack of trust in the quality of search results 
retrieved using federated search over distributed third party data. 
Hence, dataset and endpoint selection and discovery are inherent 
challenges for query distribution. These are currently hindered by the 
lack of trust-worthy and up-to-date information about the nature, 
characteristics, currentness and suitability of particular datasets for 
a given task. Given the heterogeneous and large-scale context of LOD, 
state-of-the-art semantic and keyword search techniques for structured 
data face increased query ambiguity and scalability problems already in 
single-source search scenarios. In the federated search scenarios for 
LOD, dataset selection and adoption of queries to the respective schemas 
used poses even further challenges.

As the Linked Open Data (LOD) Cloud includes data from a variety of 
domains spread across hundreds of datasets containing billions of 
entities and facts and is constantly evolving, manual assessment of 
dataset features is not feasible or sustainable, leading to brief and 
often outdated dataset metadata. That is, for instance, apparent with 
the DataHub (http://datahub.io), the largest dataset registry for open 
datasets in general and LOD in particular. Hence, given the dynamic and 
evolving nature of the LOD Cloud, particular focus should be on the 
development of scalable automated approaches, which facilitate the 
frequent assessment and profiling of large-scale datasets to enable the 
selection of suitable datasets for query federation.


The main areas of interest of this special issue include:

I) Dataset/endpoint analysis, profiling and discovery:

* query routing taking into account relevance and quality of distributed 
* dataset profile representation (vocabularies, schemas)
* novel applications and techniques for dataset profiling
* automated approaches to dataset analysis and exploration
* analysis/monitoring of dataset and graph dynamics
* topic profiling of datasets
* dataset and graph preservation
* assessment of dataset evolution
* assessment of dataset schema conformance and evolution
* dataset quality analysis for query routing
* assessment of data evolution and change propagation
* automatic and semi-automatic data linking

II) Distributed semantic search:

* federated search for Linked Data
* semantic annotation and expansion for keyword queries
* keyword query interpretation and disambiguation for Linked Data
* graph-based keyword search
* multilingual search
* categorization of keyword query interpretations and results, faceted 
* fusing, cleaning, ranking and refining search results
* scalability & performance of distributed data queries
* novel applications for federated search over Linked Data


Submissions to this special issue should follow the journal's 
guidelines. Initial submission should be made as a PDF document via 
Easychair at https://easychair.org/conferences/?conf=ijswisdp2014. 
Please include, in accordance with the journal's guidelines, 5-10 
keywords, to assist both with indexing and reviewer selection.

If a submission is based on a prior conference or workshop publication, 
the journal submission must demonstrate substantial 
differences/improvement. For such an extended manuscript, please 
explicitly cite previous publications on which the current submission is 
based, along with a note explaining the main differences/improvements.


Submission deadline: 8 December 2014
Notification: 8 April 2015
Revisions due: 22 May 2015
Estimated publication: Q3/Q4 2015


Elena Demidova, L3S Research Center, Germany
Stefan Dietze, L3S Research Center, Germany
Julian Szymański, Gdańsk University of Technology, Poland
John Breslin, NUI Galway, Ireland

Contact for further information: ijswisdp2014@easychair.org.
Received on Tuesday, 26 August 2014 10:37:33 UTC

This archive was generated by hypermail 2.4.0 : Tuesday, 5 July 2022 08:45:38 UTC