Data Science Festival Day 2 - Lightning talks at Skyscanner

Data Science Festival Skyscanner Event (Ballot ticket only)

Event near Moorgate Station

THIS IS NOT AN EVENT TICKET REGISTERING ON MEETUP.COM DOES NOT GUARANTEE YOU A PLACE AT THIS EVENT.

Please register for a ballot ticket here: https://www.eventbrite.co.uk/e/ballot-ticket-data-science-festival-day-2-lightening-talks-at-skyscanner-tickets-32952301220

Due to the popularity of Data Science Festival events, we are now allocating event tickets via a random ballot. Registering enters you into the ticket ballot for the Data Science Festival Event at Skyscanner on Tuesday 25th April 2017, the ballot will be drawn on the 18th April 2017. Those randomly selected will then be e-mailed tickets for the event, with the joining details

The Data Science Festival is the first of its kind as the only community led, free to attend Data Science Festival in the UK. The Skyscanner event will be an evening of lightning talks with thought leading Data Scientists.

18:00 Attendees start arriving

18:15 Welcome & let's get started (Bryan Dove, Skyscanner)

18:20 Lightning talks (10 mins each):

Mike Passey, Neal Lathia Skyscanner, Fabrizio Silvestri Facebook, Dima Karamshuk Skyscanner.

19:05 Break

19:15 Lightning talks (10 mins each):

Ruth García Skyscanner, Changtao Zhong Twitter, Arthur Vivian Skyscanner, Ryszard Kaleta Twizoo.

19:55-20:30 Networking & drinks

20:30 Close

Winning with Data Science at Facebook (Mike)

Data science means lots of different things to different people. In this talk I’ll describe how we think about impacting products with data science at Facebook.

On the Use of Embeddings in Search (Fabrizio)

One of the main innovations rediscovered in the last years in search and machine learning is the concept of Embeddings. In search, embeddings have been used in many different applications including retrieval, advertising, and recommender systems. In this talk we are going to show some applications of vector space embeddings that have considerably improved the state of the art. All the applications shown have been adopted by main search companies in production in their systems.

How Different are Your Different Social Network Personae? (Changtao)

This talk investigates when users create profiles in different social networks, whether they are redundant expressions of the same persona, or they are adapted to each platform. Using the personal webpages of 116,998 users on About.me, we identify and extract matched user profiles on several major social networks including Facebook, Twitter, LinkedIn, and Instagram. We find evidence for distinct site-specific norms, such as differences in the language used in the text of the profile self-description, and the kind of picture used as profile image. However, different genders and age groups adapt their behaviour differently from each other, and these differences are, in general, consistent across different platforms.

Bootstrapping a Destination Recommendation Engine (Neal)

In many different web services, we hear about machine learning for recommendation systems that help users tackle information overload - there are simply too many movies, songs, and books for users to usefully browse through. Travel is a little bit different - the world does not have millions of cities - but finding new, interesting places to travel to is still a challenge. Years ago, Skyscanner started it’s ‘everywhere’ search, allowing users to find the cheapest places to travel to. Since then, research has demonstrated that price is one of many factors that make a place attractive. In this talk, I’ll discuss how we’ve bootstrapped a destination recommender system using the rich implicit data generated by Skyscanner’s millions of users, simple algorithmic approaches, and experiments that gauge how localised and personalised recommendation affects user engagement.

Modern Advertising: Assessing the Quality of Ads Beyond Only Clicks (Ruth)

Assessing online advertising user experience is important to avoid driving users away from your site and to provide feedback to advertisers on the quality of their ads. Skyscanner is increasing their advertising inventory, adding modern ads and removing older advertisement. For example, in 2016 “Inline ads” were added on the first position of the flight day view page. In this talk, I’ll will discuss my experience in assessing the quality of ads in Skyscanner beyond only clicks and my journey in enforcing the proper tracking of user experience before, during and after an ad has been clicked.

Data Mining for Optimizing Content Caching and Distribution (Dima)

As a leading travel marketplace, Skyscanner is serving a daily load of up to a dozen billion flight itineraries to its users across the globe. The distribution of travel quotes at such a scale requires caching mechanisms optimized for minimizing the load on the partners (airlines and travel agencies) and maximizing the relevance and comprehensiveness of the itineraries to the travelers. This talk will be focused on using data mining approaches for optimizing dynamic content distribution at scale and will shed the light on Skyscanner's efforts in this direction.

Enabling Log Everything (Arthur)

We believe that data should be at the heart of every decision at Skyscanner, so it’s important that our engineers have the tools to seamlessly log the data that will help them with those decisions. In this talk, we discuss the approach we’ve taken to enable this and reflect on some of the challenges and lessons learnt. We'll briefly look at some of the technology underpinning this, including Kafka, Logstash, Elasticsearch, Secor, AWS (S3, lambda), Samza, Protocol Buffers and others, and look at the custom tooling built in Skyscanner.

Rapid Machine Learning Development with Social Data (Ryszard)

In this talk I will introduce the unique requirements and exciting challenges presented by mining social media for high-quality, user-generated content. I will outline how we approach these both as data scientists and big data engineers, finally sharing first-hand experience of building tooling and architecture that supports rapid model development.

Check in at http://www.datasciencefestival.com to find out more details as they become available. THIS IS NOT AN EVENT TICKET