{practical.magic}

In recent years, community Question Answering forums such as StackOverflow, Quora, and BG-Mamma have gained a lot of popularity as a source of knowledge. These forums typically organize their content in the form of multiple topic-oriented question–answer threads, where a question posted by a user in the past is followed by a possibly very long list of other users' comments intended to answer the question. Many such on-line forums are not moderated, which often results in noisy and redundant content, as users tend to deviate from the question and start asking new questions or engage in conversations, fights, etc. Yet, they represent a rich source of information, which can help answer a number of new questions, as people often ask similar things again and again. I will explore three general problems related to such forums: (i) deciding which answers are good, (ii) finding related/duplicated questions, and (iii) finding good answers to a new question. This will involve models based on deep learning and semantic/syntactic kernels. Part of this work was integrated in a production system, e.g., as thumbs up in a forum for (i), or as part of a smart search for (iii). I will also introduce some promising extensions of this work in directions such as application to Arabic (can we apply these models to Arabic medical forums) and Bulgarian (how about the BG-Mamma forum?), cross-language question answering (can we answer a question in Arabic using a forum that is in English), fact checking (many answers in the forum look superficially good, but which of them are actually factual), trollness detection (can we find the forum trolls and their answers), answer justification (can we find and group contradictory answers), and interactive cQA (can we turn an entire forum into a chatbot, e.g., one that can be integrated in Alexa). This research was performed by the Arabic Language Technologies (ALT) group at the Qatar Computing Research Institute, BHKU. It is part of the Interactive sYstems for Answer Search (IYAS) project, which is developed in collaboration with MIT-CSAIL.

AI-powered projects come with a number of challenges on how work is organized. This talk explores the specifics of developing a Machine Learning project:

• How do we plan Machine Learning development;
• How do we test Machine Learning components;
• How do we track model versions;
• How do we debug Machine Learning components.

With software projects that are being developed by multiple teams of numerous people, daily releases sound quite scary. Naturally, this shouldn’t be the case. We will share our experience on developing an e-commerce platform of a major sports brand. What is the development process that allows us to deliver working software to millions of users on a daily basis? We will share how we test the software and what tools help us, so that we don’t break the output cadence.

React, Angular, and Vue are the three most popular front-end frameworks at the moment. In this talk we will take a look at the problems that they solve and why they make things so much easier for us. We will see how each of them implements specific concepts such as Dom Rendering/Virtual Dom and Change Detection. We will try to learn something new from each framework, without trying to determine which one is ‘best’.

A deep headfirst dive into the murky swamp of finding out what kind of people make the things we use all the time, on and offline. The concept of design will be our guiding light in the swamp. And just before we drown together in the depths, we will devote our last breaths to important considerations about the irrational in our choices. Let's see what's in the mind of a designer.

Authorization by itself is a simple task - give a bunch of users a bunch of rights for a bunch of objects, right? In this talk we will share lessons learned from 7 years of experiments with various approaches and their limits. We will learn why we got back to the legacy option of server session and why we were forced to write an authorization library of our own in JavaScript.

Minimalism in art arises as a protest against the aspiration of artists to express themselves via peculiar unintelligible and unconventional means. Minimalists instead use little but sufficient magic. To what extent can these ideas be applied to finding balance between theoretical elegance and practical usability while creating abstractions and implementing them via program code? How to reduce the initial shock that our code causes in others? This presentation will give guidance to the answers of these questions.

This talk will bridge across different aspects of human knowledge: Literature, Philosophy and Artificial Intelligence. We will try to answer the question what does it mean for an AI to inhabit a world and whether this sense of inhabiting can be defined and understood. We will concentrate on neural networks of the “encoder” type, looked at through the lens of how they learn to make copies of objects and phenomena they have witnessed, creating a descriptive language of their own in the process. Every reader of fantasy and science fiction can tell whether they are in a world crafted by Tolkien, George Martin or Lucas, but how does an AI understand that? In this talk we will try together to discover the “magic of literature” in a most practical sense, by means of machine learning models – a useful task for every writer, content designer or game developer.

This talk is a brief introduction to virtual reality, answering the following questions:

• How do we place UI in virtual worlds?
• How do we interact with the UI and the virtual objects?
• What are the specialized interaction controllers?
• How can we include voice recognition in virtual worlds?

Statistics is a very powerful instrument. It is at the core of machine learning - a trending technique that produces impressive results. It can also be used to prove the correlation between the birth rate and the number of storks in a country. How can we misuse statistics to prove our point? How does machine learning 'learn' bullshit when we don't take into account how statistics works?