KDD 2015 Recap; Data Down Under
Our resident scientist, Peiman Barnaghi has just returned from a trip down under for KDD ‘2015 in Sydney, Australia. The annual conference, on Knowledge Discovery and Data Mining (KDD), is the premier international forum for Data Science, Data Mining, Knowledge Discovery and Big Data. The conference brings together practitioners, researchers, academia and industry professionals to share knowledge, ideas and findings in a highly educational environment.
The conference supported by the likes of Google, Yahoo, Alibaba, Baidu and Facebook featured 4 Keynote presentations, 11 talks, a whopping 228 paper presentations, tutorials and workshops as well as poster sessions by researchers, all spread across 4 floors of the Hilton in Sydney’s CBD.
Given the fantastic opportunity for learning and exposure, Peiman was delighted to be invited along to KDD ‘15 to present a recent paper he completed “Text Analysis and Sentiment Polarity on FIFA World Cup 2014 Tweets” as part of a workshop, entitled “Large-Scale Sports Analytics”.
Peiman’s paper aims to examine the effectiveness of a machine learning method for providing positive or negative sentiment on tweets and to find correlations between such sentiment and real-life events such as a team scoring in a game of football, which is a part of Peiman’s ongoing research in Sentiment Analysis on streaming feeds. You can download Peiman’s publication here.
Peiman’s top picks from the various talks, tutorials and research sessions are listed below:
Online Controlled Experiments: Lessons from Running A/B/n Tests for 12 years by Ronni Kohavi, General Manager, Analysis and Experimentation, Microsoft (Download Slides)
Web Personalisation and Recommender Systems, delivered by Shlomo Berkovsky and Jill Freyne from CSIRO. (Download Slides)
Automatic Entity Recognition and Typing from Massive Text Corpora – A Phrase and Network Mining Approach. (More Info and Slides)
Each research sessions dealt with ~5 topics delivered by various speakers, which are all listed at the links below.
- Topic Models and Tensors (Link)
- Big Data (Link)
- Knowledge Discovery (Link)
- Sampling and Streams (Link)