NLP on social media

Social media is a gold mine for marketing and social science investigations (among many others). The December meetup is dedicated for NLP on social media texts, i.e. we can discuss what kind of tasks a data science company faces and what kind of insights can we gain from social media texts?

Talks:

• Markus Frise,BlackSwan (http://www.blackswan.com/): Value from social text data

Brands care what their customers say online. But they struggle to collect this data and process it, so that it is useful to them. I’m going to discuss how we help Brands do this: create value from text data. Should be fun.

• Richárd Farkas, BlackSwan (http://www.blackswan.com/): NLP on social media, beyond sentiment

"Social media data has unlimited value & near limitless application". I'm introducing two business cases on social media text NLP others than the well-known sentiment analysis.

• Gábor Vattay and Eszter Bokányi, ELTE (http://complex.elte.hu/): Language and demographics: Get to know your customers

How does our demographics shape what we say? Can we figure out the demography of our customers from what they say? By identifying and quantifying the most relevant language features of the United States on Twitter, we try to answer these questions. Through analyzing the geographical distribution of the most distinctive word patterns, we show that the main socio-economical driving forces of word choice are race, religion, and the population density. On the example of Walmart, we show how we can reverse-engineer the demography of customers from their language.

Schedule:

18:30- pizza and beer, small talk with friends

19:00- Presentations + Q&A

20:00- Socializing

Speakers:

Markus Frise lives in North London. His background is in the physical sciences, demand forecasting and data analysis generally. He spent 6 years running demand forecasting at Tesco before coming to Black Swan to head up the Data Science team.

Richárd Farkas has been working on machine learning-based solutions for NLP since 2003. His main research interest is on real-world NLP applications which desire deeper linguistic analysis. He is an assistant professor on AI, machine learning and NLP at the University of Szeged and a lead data scientists at Black Swan Data.

Eszter Bokanyi is a graduate student and Gábor Vattay is a professor at the Department of Physics of Complex Systems of Eötvös University (http://complex.elte.hu/). Their main research interest is finding relationships between the digital imprints of people using various communication networks and real-world phenomena.