Semantics in the Deep: Semantic Analytics for Big Data
Dr. Dimitrios A. Koutsomitropoulos
Prof. Spiridon D. Likothanassis
Aim and Scope
Recent advances in availability of information on the Internet, storage space and web generated content have paved the way for the advent of Big Data. The well-known 4 Vs (Velocity, Variety, Volume, Value) that characterize Big Data can find a match in intelligent ways for management, manipulation and value-extraction. It is widely acknowledged that the recent upheaval in AI and especially machine learning is exactly due to these advancements. The Semantic Web can offer a well-studied, although ever advancing, toolbox that can address Big Data requirements and contribute towards their meaningful analysis. Still, there are often issues that need to be tackled with like bootstrapping, efficiency and standardized business processes for semantic analytics to achieve satisfactory results. To this end, machine- and deep-learning techniques, while being considered the poor relation for years, have been shown to have considerable contributions towards Big Data analytics and to overcome Semantic Web inherent limitations.
Therefore, the aim of this workshop is to bring together researchers and practitioners to look deeper into how Semantic Web technologies can contribute towards Big Data analytics. This can be achieved either by extracting value out of these data (through reasoning), creating sustainable ontology models, offering a solid foundation for deploying learning techniques or anything in between.
Topics of Interest
Some indicative topics of interest for the workshop include, but are not limited to:
- Ontologies for big data
- Semantic applications in big data domains including open datasets, linked data, scholarly information, e-learning, economics, insurance, sensors, bioinformatics
- Reasoning approaches for knowledge extraction
- Ontology learning
- Topic modeling
- Linked Data
- NLP and word embedding
- Semantic deep learning
- Semantic lakes
- OBDA approaches for big data access
- Evaluation techniques
- Ontologies as training sets
- Ontology evolution and learning feedback
- Scalability issues
Papers reporting original and unpublished research results on the above and related topics are solicited. Authors should submit a paper with up to 10 pages (in English) in single-column Springer format following the Springer IFIP AICT format guidelines. Submissions must be in electronic form as PDF files and should be uploaded with EasyChair. Submitted papers will be peer-reviewed by at least 2 independent members of the program committee.
At least one author of each accepted paper must be registered and present the paper at the workshop. All accepted papers will be included in the conference proceedings and published in the SPRINGER IFIP AICT (Advances in Information and Communication Technology series. Extended versions of selected workshop papers may be considered for publication in international journals.