May 31, 2018
CoreValue at Open Data Science Conference
Our CTO and Head of Research and Development Lyubomyr (Sam) Senyuk has just returned from the USA where he visited the Open Data Science Conference (ODSC) held in Boston on May 1–4, 2018. The theme this year was “The Future of AI is Here.” For closer look at the conference and his experience there, read on.
Q: Why ODSC?
ODSC East 2018 was one of the largest applied data science conferences in the world. It is a cool place to learn the latest in cutting edge topics, like AI and data-related topics, tools, and languages from the best and brightest in the field. There were three focus areas: Deep Learning; Data Analytics; Data Visualization, which was a perfect match for our increased activities within our R&D department. Since we are aiming at specific data science topics in Pharma/Life Sciences, including machine learning and NLP, hearing from some of the core contributors to many open source tools, libraries, and languages was special and beneficial.
Q: What was your first impression of a conference on such a scale?
It was my first time visiting ODSC, and it is huge. An amazing conference, impressive. The sheer talent that people have, and their ability to take even the strangest ideas and invest an insane amount of time into them to come up with something that moves the work forward, is so inspirational. I felt sorry it wasn’t physically possible to visit every session or meeting. Just imagine 400 business professionals and 4,000 data scientists all in one place: 4-days of talks, tutorials, workshops, case studies, and professional trainings around the world of data and machine learning.
Q: So you found it worth visiting. What was your takeaway from the conference?
ODSC is unique in that it attracts an audience of decision-makers, such as CTOs and lead data scientists, as well as decision-influencers, that is to say the people who actually use and build analytic tools of the future. I’ve visited a dozen presentations and they all were worth it
My takeaway from the conference is that data science, AI and machine learning are becoming mainstream. Many industries use them right now in operations as diverse as screening for autism, identifying fake news, and deep learning in aerospace. Keynote Cathy O’Neil and Stephanie Kim also touched on ethical problems, discrimination, equality and false stats influenced by bias factors in Data Science and AI.
Besides specific concerns related to R&D topics, machine learning in Life Sciences presentations were great takeaways. For instance, going beyond NLP in sepsis detection by adding user characteristics and patterns from the doctors’ notes allows detection of disease much faster, even by minor slight indications. This results in effective treatment at an earlier stage.
Q: What are the future plans of our R&D department?
We are thinking about the perspectives, and it’s obvious that we are moving in the right direction by researching and building data science solutions that can become powerful instruments commonly accessible. So, CoreSearch is extending the team and developing our competencies in order to bring value both to our clients and the general public.
Presently, as the part of our wide cooperation with one of the key players of the US Pharma market, we are participating in a project that incorporates revision and upgrading their KOL (Key Opinion Leader) identification algorithms. There are several approaches to KOL identification operating with different types of data, but usually relationship network analysis is applied. We use historical prescriptions information and data from medical publications additionally. The Key Opinion Leaders are great influencers on medical practitioners, so they are important for pharma marketing objectives, enabling better insights together with more precise drugs acceptance analytics, and consequently better sales efficiency.
One more subject covered during the conference – Topic modeling – also translated to our present project. Topic modeling is a becoming a well-suited tool for big data in scalable environments, and one of the presentations at the conference exhibited the importance of its application in Healthcare and Life Science areas.
We also continue working with NLP-based models for researching PubMed, which is an archive of biomedical and life sciences journal literature at the U.S. National Institutes of Health’s National Library of Medicine. It would also assist in researching cancer publications to mine for valuable information from unstructured clinical text. This would enhance clinical practice through improving models for disease progress, preventing over-treatment, and narrowing down to the cure.
So if you have any ideas and care to join our team, you are welcome to drop me a line at firstname.lastname@example.org