The ESWC 2018 and My PhD Symposium Paper – a Review

Share :

 

34671478_10215516073511027_3575295983107440640_o

In the first of June 2018, I participated the 15th Extended (or European) Semantic Web Conference (ESWC) 2018 held in Crete, Greece. That was my first time conference ever where I presented my paper in front of the semantic web community. ESWC by itself is one of the major venue for discussing the latest scientific results and technology innovations around semantic technology. Luckily, I had an opportunity to present my accepted paper on the ESWC PhD Symposium Session, a part of ESWC conference event.

Overview

ESWC 2018 was held in June 3rd until 7th of June in the Aldemar Knossos Royal Village Conference Center, in Heraklion, Crete, Greece. The venue basically was a great conference hotel which was very nice and exciting because it’s just located closely to the Anissaras Beach, Crete.

According to the ESWC homepage. ESWC 2018 has accepted around 31 papers for Research Track, 8 paper for In-use track, 9 papers for Resource Track and some papers including Workshops & Tutorial, PhD Symposium and Poster papers. It was managed to 8 consecutive days with 2 days for PhD Symposium, Workshop & Tutorials and 3 days for the main conference, challenge and poster session.

The PhD Symposium Session

34473262_10215508697726637_4621212857185337344_n

Regarding the PhD Symposium, the presentation was started by the interesting keynote from Dr. Annalisa Gentile, a researcher from IBM Inc., who talked about a Hitchhiker’s Guide to PhD. This talk was not only nice and but also inspiring. It gave us a good point of view and lesson learned on how to conduct the PhD research in an appropriate way especially for linked data and semantic web researcher. She also gave us a lot of tips and tricks to catch up our PhD research successfully.

I got my turn for presenting my paper on the first day of the conference along with the other participants of the ESWC’s PhD symposium. Actually, there should be ten presentations that could be presented at that moment, but unfortunately one participant was disappear, so there were only nine presentations playing around. Regarding the PhD Symposium track, it’s accepted papers topic were quite interesting including: The quality of data interlinking, ontology evolution framework, the fusion of semantic web and machine learning etc. All of the papers have been officially published now and it can be found here. After the presentation, I got a lot of valuable feedbacks from other researchers which gave me some input for ideas, insight and also suggestion to strengthen my research topic and alternative options which could be taken to support answering my research question.

My Research Topic

My PhD Symposium paper was titled Semantic Query Federation for Scalable Security Log Analysis”. The motivation of the research is coming out based on the problem of the emergence of sophisticated, non-trivial, variety of cybersecurity attacks as they lead to multiply the difficulty and the complexity of keeping the system safe. Most organizations are now threated serious loss from these attack. Therefore, it is necessary to analyze a wide range of log files that potentially provide clues about malicious activity. However, log data is typically unstructured, big, heterogeneous and distributed. Existing security analysis approach (e.g. SIEM, IDS etc.) mainly focus on regular expressions and statistical indicators and do not directly provide actionable insight. Manual activity of tracing an attack by linking distributed log events remains tedious and slow of response.

My research question is “How to identify causally related security events from disparate log information in a scalable manner?”. This question is important since the problem is real and happening but it has not been addressed yet so far. The answer of the question will have a huge impact regarding the new approach to overcome the cybersecurity threat efficiently in scalable manner. It also can improve the security awareness, accelerating incident response and therefore it can prevent organizations from serious loss because of the cyberattack.

We propose an innovative approach of scalable security log analysis that leverage linked data technology. Linked data is a method of publishing structured data so that it can be interlinked and become more useful through semantic queries (Wikipedia). Linked data technology provide mechanism to retrieve data from distributed dataset using semantic query federation. It also provide capability to retrieve unseen information from existing graph data via semantic (stream) reasoning that will hence allow us to detect potential attacks and suspicious behavior in real-time.

To this end, we have already developed the generic log vocabulary, lifting raw log data from different log sources (our case Linux Sys-log, Authentication-log, Apache-log, Firewall-log, Windows Event-log) and transform them into RDF data (JSON-LD in our case). We also enriched and linked the RDF log data against the background knowledge so it can be used to derive more information. For the future work we be to extract event from the RDF data and integrate them.

The Workshops & Tutorials

In the second day of ESWC Conference, I attended several workshop-track presentations such as presentation from Dr. Ghislain, a researcher from Mondeca, who talked about benchmarking commercial RDF stores with publication office’s dataset. This is interesting, as the presentation showed a performance comparison between commercial RDF stores such as Virtuoso, GraphDB, Neo4j, Stardog, Oracle, lazegraph, Marklogic etc. Later on, we can determine which commercial RDF store that most suitable for a certain case, according to the architecture and infrastructure settings. There were also another interesting workshop such as a workshop that more focus on the intersection between deep learning and semantic web. It discussed more on how to create an intelligent artifacts that emulate human capacities such as reasoning, validating, and predicting.

I also attended some interesting tutorials such as a tutorial delivered by Dr. Antoine Zimmermann, who talked about semantic interoperability and integration among heterogeneous data source. He introduced SPARQL-Generate, which was an extension of SPARQL for query not only RDF dataset, but also documents in arbitrary formats. He demonstrated how this extension could generate RDF-Graph from JSON document using SPARQL Template Transformation Language (STTL).

The Main Conference

The main conference of ESWC 2018 was started on the third day, from 5th – 7th of June 2018. It was the moment when the main accepted papers were presented and everybody had been waiting for. The conference paper was distinguished by three different tracks i.e. research track, in-use track, and resource track. The research track basically covered research that more focus on the problems and areas that directly impact the development of the Semantic Web. The in-use track exposed research that focused more on the exploitation of semantic web outside the lab, the use of semantic web to address the real-life problem and its potential using in industry on a broad sector. The resource track exposed the development of resource that provide the necessary scaffolding to support the scientific publications such as dataset, ontologies and vocabularies, software framework, tools, library, APIs etc.

It included several interesting topics such as the intersection of the Semantic Web with Machine Learning, Natural Language Programming (NLP), Social Web, Ontologies and Vocabularies, Data Management and Querying, Knowledge Graph, Streams and Reasoning, Semantic Search etc.

Poster Session

WhatsApp Image 2018-06-07 at 9.38.42 AM

In the last day (7th June) of ESWC Conference, all participant of PhD Symposium had to exhibit their research poster on the Poster Session together with the participant from Poster track. Here we got a chance to present and explain our idea among the visitors which were mostly scientist, experts and other researcher from the semantic web community. Later on, they could take a look closer to the poster and gave some feedbacks to the presenter. I got some interesting questions the feedbacks here as well.

 

 

 

 

Conclusion

Overall the ESWC 2018 conference was incredible. I was so happy to get my paper accepted here as it was my first conference ever in the international forum of semantic web community. Meeting a lot of influenced researcher all over the word gave me a high motivation to work hard and hopefully can contribute knowledge in the area of semantic web. This conference gave me the opportunity to engage with semantic web community, having an interesting talk and discussion with experts from the semantic web. And the most important thing is that, this event motivated me so much deeply learn about semantic web. I hope to have another chance to submit my next paper to this conference or the other primary semantic web conference e.g. International Semantic Web Conference (ISWC), Semantics, EKAW etc. The next ESWC Conference (2019) will be held in Portoroz, Slovenia. I recommended everyone including PhD student who working in the area of semantic web, knowledge graph, machine learning and soon to submit and present their work as it give a valuable input of idea, new insight and the opportunity to develop networking with other semantic web researcher.

This entry was posted on Friday, August 24th, 2018 at 10:41 am and is filed under Uncategorized. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

Leave a Reply

CommentLuv badge