Understanding decisions, relationships and behaviors through text… and Natural Language Processing
Event details
| Date | 11.09.2018 |
| Hour | 10:15 › 11:30 |
| Speaker | Léa Deleris |
| Location | |
| Category | Conferences - Seminars |
Abstract:
Have you ever been stuck in meetings with discussions going in circles, meetings from which no decision emerges or one with which everyone is frustrated?
Do you sometimes feel that you misunderstood the relationships among members of your professional or personal circles?
Have you ever wondered how you could get to stop smoking, eat healthy and start exercising?
If you answered yes to any of those questions, come to this talk where I will present three projects that aim at understanding decisions, relationships and behaviors (respectively) drawing from text sources.
In the DECODE project (Decision Conversations Decoded and Explained), we analyze transcripts of discussions, typically meetings, in which a decision is being discussed. Our goal is simple: follow the decision process, automatically keep track of the options that are being considered and why they are being proposed and summarize the information in an actionable form.
In the Relationship Insights project, we seek to decipher how we relate to one another. Making use of case notes, speech transcripts, posts, books etc. as the source of information about interactions, we extract qualitative and quantitative information about interactions among persons that extends beyond building a social network. Such information is relevant for instance in social care, sales and intelligence work.
In the Human Behaviour-Change Project (HBCP), a collaboration between University College London, Cambridge University, University of Aberdeen and IBM Research, we are developing a knowledge system that automatically extracts and reasons with findings from behavior change interventions studies, starting with smoking cessation interventions. Our ambition is to create a system that can help researcher and policy makers better understand what works and in which context?
Bio:
Léa received M.S. degrees in Economic Systems from Ecole Polytechnique (France) in 1999 and in Management Science and Engineering from Stanford University in 2001. She joined IBM Research in 2006 just after graduating from Stanford University with a PhD. in Management Science and Engineering.
She has been a Research Staff Member and Manager in IBM Ireland since April 2010. One of her previous roles was to lead the Risk Management Collaboratory project (see Projects). Before moving to Ireland, she was a member of the Risk Analytics group of the Business Application and Mathematical Science Department at the IBM TJ Watson Research Center.
Léa currently manages a team of 10+ researchers and software engineers working on a diversity of projects that seek to make use of artificial intelligence and natural language processing to support decision making in a variety of situations including social care, health care, chemistry but also sales management.
Her personal research interests lies in decision theory and risk analysis on one side and artificial intelligence (and in particular natural language processing) on the other side. Currently, her work presents a mixture of applied and conceptual work, including (i) information extraction in the field of behavior change (ii) causal modeling for recommendation of interventions in smoking cessation, (iii) learning algorithms for Bayesian networks from messy data and (iv) targeted sentiment analysis.
Host: Swiss Data Science Center, Olivier Verscheure
Have you ever been stuck in meetings with discussions going in circles, meetings from which no decision emerges or one with which everyone is frustrated?
Do you sometimes feel that you misunderstood the relationships among members of your professional or personal circles?
Have you ever wondered how you could get to stop smoking, eat healthy and start exercising?
If you answered yes to any of those questions, come to this talk where I will present three projects that aim at understanding decisions, relationships and behaviors (respectively) drawing from text sources.
In the DECODE project (Decision Conversations Decoded and Explained), we analyze transcripts of discussions, typically meetings, in which a decision is being discussed. Our goal is simple: follow the decision process, automatically keep track of the options that are being considered and why they are being proposed and summarize the information in an actionable form.
In the Relationship Insights project, we seek to decipher how we relate to one another. Making use of case notes, speech transcripts, posts, books etc. as the source of information about interactions, we extract qualitative and quantitative information about interactions among persons that extends beyond building a social network. Such information is relevant for instance in social care, sales and intelligence work.
In the Human Behaviour-Change Project (HBCP), a collaboration between University College London, Cambridge University, University of Aberdeen and IBM Research, we are developing a knowledge system that automatically extracts and reasons with findings from behavior change interventions studies, starting with smoking cessation interventions. Our ambition is to create a system that can help researcher and policy makers better understand what works and in which context?
Bio:
Léa received M.S. degrees in Economic Systems from Ecole Polytechnique (France) in 1999 and in Management Science and Engineering from Stanford University in 2001. She joined IBM Research in 2006 just after graduating from Stanford University with a PhD. in Management Science and Engineering.
She has been a Research Staff Member and Manager in IBM Ireland since April 2010. One of her previous roles was to lead the Risk Management Collaboratory project (see Projects). Before moving to Ireland, she was a member of the Risk Analytics group of the Business Application and Mathematical Science Department at the IBM TJ Watson Research Center.
Léa currently manages a team of 10+ researchers and software engineers working on a diversity of projects that seek to make use of artificial intelligence and natural language processing to support decision making in a variety of situations including social care, health care, chemistry but also sales management.
Her personal research interests lies in decision theory and risk analysis on one side and artificial intelligence (and in particular natural language processing) on the other side. Currently, her work presents a mixture of applied and conceptual work, including (i) information extraction in the field of behavior change (ii) causal modeling for recommendation of interventions in smoking cessation, (iii) learning algorithms for Bayesian networks from messy data and (iv) targeted sentiment analysis.
Host: Swiss Data Science Center, Olivier Verscheure
Practical information
- General public
- Free
Organizer
- Swiss Data Science Center, Olivier Verscheure
Contact
- Cindy Ravey