IMPORTANT DATES
• Deadline for submission: June 26, 2022
• Notification of accepted works: July 25, 2022
• Final version for publication in the proceedings: July 31, 2022 (firm deadline!)
DSW – Dataset Showcase Workshop
The publication and availability of datasets (open or not) have become highly relevant due to the great attention given by various segments, such as the media, industry, academia and government. The task of making data available is important for numerous reasons, ranging from its reuse in digital applications developed by society, to the possibility of reproducing experiments and developments by the scientific community. Therefore, in the context of the Brazilian Database community, making datasets available is intrinsically important due to the promotion of new research and development issues that such a task can achieve.
Thus, the purpose of SBBD DSW is to provide a forum to share and discuss the ways of building and organizing data sets that serve as a basis for the research work carried out in the Brazilian scientific community. The contribution of an article to be published in SBBD DSW is the final product in the form of a dataset, usually extracted from some database or Web platform, cleaned and treated, often augmented with external data, and possible to be reused in other scenarios, or for reproducing experiments. The contribution is the product in the form of the dataset, but the article must present all the information necessary to understand and use it.
Submissions
Articles must describe the data as handled by their authoring team and made publicly available. In principle, the dataset needs to be useful and quickly reused by third parties, through, for example: adding value to the data for other people in the community, being pre-processed or filtered in some way; easy-to-understand organization through a schema, dictionary, taxonomy, ontology, or other formalism; access facilitated by its own mechanisms; or presentation of distinguishable quality through complex curation and cleaning.
Articles submitted to the SBBD DSW must be written in Portuguese, English or Spanish, must have an abstract in English, be in the SBC format, be between 6 and 10 pages long (with up to two extra pages being possible just for references and acknowledgments) and be submitted via JEMS (available soon). Each submission should include as appropriate and in the order the authoring team prefers:
- description of the data source(s) and the complete methodology for collecting or generating them (with public availability of the tool used to create or generate the data, if applicable, and);
- description of the storage mechanism, including, if any, schema or data dictionary, taxonomy, ontology, or other formalism that facilitates its use by third parties;
- quantitative description of the database created, as well as initial analysis of the data characterizing the database and reporting the amount of missing data, tables and other important information that can be used to assess the quality of the dataset;
- description of how the data have been used (if any, refer to articles published or in the process of submission that use the data and how) and of its unpublished character, because even with the use in any article or submission, the complete description of the dataset according to performed in the submission to the DSW must be unpublished;
- discussion of existing challenges and possible limitations in the use of data;
- ideas for different uses of the data, scenarios, research questions that could be elaborated or solved from the availability of the data, and what improvements could be applied to the data;
- and public location for download, as the dataset must be publicly available at the time of submission of the article for review. Preferably, use specialized online platforms suitable for public data maintenance, including github, zenodo, figshare or osf (i.e. avoid private repositories such as Google Drive directory, dropbox, onedrive and the like).
It is important to note that submissions to the SBBD DSW should not overlap with submissions to the main event and its satellites, as the contribution is completely different from the articles currently published in the SBBD. In particular, data showcase articles are *not*:
- survey, systematic reviews, empirical studies or experimental evaluation;
- articles with proposed tools for data generation and processing;
- based on weak and doubtful collection heuristics; or
- simple applications of generic tools to generate data that can be quickly and easily used by anyone.
Topics of Interest
The topics of interest for the SBBD DSW are the same as the main conference, expanding to current research topics in various areas of Computer Science and other sciences, as well as different contexts related to government, education, culture, economy, transport and health. In particular, articles are expected to present datasets that can be used in research related to (non-exhaustive list):
Scientific Applications, Data Science and Interdisciplinary with other Sciences, including e-science;
Applications and Areas Related to Databases (data analysis and visualization, machine learning, digital libraries, data mining, information retrieval, social networks, recommender systems, information systems, Web, workflow and the like);
Other areas related to Computing (including, but not limited to, all those that have an Interest Group or Special Commission at SBC, and their applications such as benchmarks, baselines, ground truths and the like);
Different types of Databases (active, Web, streams, strings, documents, in the cloud, interconnected data, Semantic Web and RDF, heterogeneous, semi-structured, XML, mobile, sensors, multidimensional, temporal, spatial and GPS, multimedia, NoSQL , NewSQL, statistics and the like);
Data Engineering (data warehouses and OLAP; authorization, privacy, anonymization and security in databases; information integration and interoperability; data processing on new hardware; data provenance).
Chairs of DSW
Carina F. Dorneles, UFSC
Mirella M. Moro, UFMG
Eric Araujo, UFLA
Previous editions are available at http://bit.ly/sbbd-dsw