Fwd: Network Research Seminar: Luca Aiello - Ten Dimensions of Social Relationships - 28th May
2 views
Skip to first unread message
Marc Santolini
May 27, 2020, 4:17:15 AM5/27/20
to network-se...@googlegroups.com, Liubov Tupikina
Dear colleagues
Social relationships are the key determinant of crucial societal outcomes, including diffusion of innovation, productivity, happiness, and life expectancy. To better attain such outcomes at scale, it is therefore paramount to have technologies that can effectively capture the type of social relationships from digital data. NLP researchers have tried to do so from conversational text but mostly focusing on sentiment or topic mining, techniques that fall short on either conciseness or exhaustiveness. We propose a theoretical model of 10 dimensions of social relationships that is backed by decades of research in social sciences and that captures most of common relationship types. We trained a deep-learning model to classify text along these ten dimensions, and we reached performance up to 0.98 AUC. By applying this tool on large-scale conversational data, we show that the combination of the predicted dimensions suggests both the types of relationships people entertain and the types of real-world communities they shape. As a use case, we show how this technology can be used in the domain of employer care by predicting Nokia employee satisfaction from online review text.
It is with pleasure that we will have the first network seminar speaker after a long break!
We will hold this online network seminar on Zoom this Thursday 28th May at 5pm-6pm.
Luca Maria Aiello, a researcher from Bell labs Cambridge team, will give a talk about ten Dimensions of Social Relationships.
There is a registration page so that everyone can get a link to the online meeting. https://events.cri-paris.org/e/886
There is a registration page so that everyone can get a link to the online meeting. https://events.cri-paris.org/e/886
Ten Dimensions of Social Relationships
Social relationships are the key determinant of crucial societal outcomes, including diffusion of innovation, productivity, happiness, and life expectancy. To better attain such outcomes at scale, it is therefore paramount to have technologies that can effectively capture the type of social relationships from digital data. NLP researchers have tried to do so from conversational text but mostly focusing on sentiment or topic mining, techniques that fall short on either conciseness or exhaustiveness. We propose a theoretical model of 10 dimensions of social relationships that is backed by decades of research in social sciences and that captures most of common relationship types. We trained a deep-learning model to classify text along these ten dimensions, and we reached performance up to 0.98 AUC. By applying this tool on large-scale conversational data, we show that the combination of the predicted dimensions suggests both the types of relationships people entertain and the types of real-world communities they shape. As a use case, we show how this technology can be used in the domain of employer care by predicting Nokia employee satisfaction from online review text.
Bio: Luca Maria Aiello is a Senior Research Scientist in the Social Dynamics team at Nokia Bell Labs Cambridge, UK. Formerly, he has been a Research Scientist at Yahoo Labs for almost 5 years and a research Fellow of the ISI Foundation in Torino, Italy. He conducts interdisciplinary research in computational social science and network science. His work has been covered by hundreds of news articles published by news outlets worldwide including Wired, WSJ, and BBC. He is currently working on NLP research to model fundamental aspects of human relationships from conversational text.
Marc and Liubov
Take care and join us for our Network seminar
Reply all
Reply to author
Forward
0 new messages