What I saw and liked at K-CAP 2019
5 months ago by Hamada Zahera
It has been a great experience attending the 10th Knowledge Capture (KCAP) conference as one of the top-tier conferences in knowledge representation and semantics computing. This year, KCAP2019 was in Marina del Rey, California. It received 104 valid submissions (89 long papers, 15 short papers). Our research group (DICE) had 5 accepted papers from 11 submissions.
On the first day of the conference (19th November), there were different tutorials and workshops related to Natural Language Processing (NLP) and hybrid questions answering. I enjoyed the presentations by Mr. Jose Manuel about hybrid techniques for knowledge-based NLP. It gave me insights about incorporating pre-trained language models and external knowledge graphs in text processing.
In the evening, I was happy to present my paper “Jointly Learning from Social Media and Environmental Data for Typhoon Intensity Prediction”, about the research I completed in my first-year PhD. In particular, we explore how to leverage social media posts to improve disaster operations during emergencies.
Photo1: Hamada is presenting his poster “Jointly Learning from Social Media and Environmental Data for Typhoon Intensity Prediction”
The second day started with the session “Towards Machines that Capture and Reason with Science” by Mr. Peter Clark from Allen Institute for AI. Mr. Peter gave us a general overview about the research at the Institute.
In the afternoon session, I presented an accepted research track paper of my college Abdullah F. Ahmed (who couldn’t attend the conference, because of his USA visa rejection). The title of his paper is “Do your Resources Sound Similar? On the Impact of Using Phonetic Similarity in Link Discovery”. The authors study the impact of phonetics similarity in discovering links among resources in different RDF datasets.
Photo2: Hamada is presenting Abduallah et.al paper “Do you Resources Sound Similar? On the impact of Using Phonetic Similarity in Link Discovery”
On the last day of the conference, I had the opportunity to meet researchers from different institutes, and discussed with them some interesting papers presented at the conference.
KCAP2019 was the first conference I have ever attended. Personally, I think it was well organized and very successful. I fully enjoyed and learned a lot from the presented tutorials and research papers. KCAP2019 helped me to focus on further research directions, especially in knowledge graphs (construction, completion and transfer learning).