Amar Viswanathan
ML Manager/ Researcher/ Research Scientist
I am a Machine Learning Research Manager at Verisk AI. As a researcher I work on a gamut of problems that fall at the intersection of Knowledge graph representation and construction, optimization, vision and language. My day to day work encompasses three fields of AI :
- Language : NLP, deep learning for information extraction, large scale question answering.
- Vision & Language : Visual question answering, Visual query reformulation.
- Knowledge Graphs : Multi modal knowledge graph construction, representation and optimization.
Prior to Verisk AI, I was a Lead ML Scientist at Best Buy AI where I worked on creating large scale knowledge graphs for the search and recommendation platform. Prior to that I was a Staff Research Scientist at Siemens Corporate Research. I got my PhD at RPI, where I worked with Jim Hendler and Deborah McGuinness on knowledge graphs, query reformulation, entity summarization, knowledge graph construction, Ontology development, Ontology alignment. I also had a chance to work on the IBM Watson question answering framework at RPI.
I had two fruitful internships at IBM TJ Watson research center and worked on summarization and query reformulation.
Before RPI I also worked as a senior software engineer in text analytics at Infosys Technologies limited where I focused on Natural Language processing, specifically pattern oriented NLP.
For more of my open source contributions : Follow @amarviswanathan
News [Archive]
Oct 22, 2020 | Presented our work on Multi modal scientific knowledge graphs at CIKM 2020 |
---|---|
Oct 22, 2020 | Chaired the CIKM Demos and Posters session with Rajiv Ranjan and Zhenyu Wen |
Recent papers [Full list]
- CIKM Knowledge Graphs IE NLPA multi modal knowledge graph for deep learning papers and codeIn Proceedings of the 29th ACM International Conference on Information and Knowledge Management 2020