We are living in a society where change is exponential - perhaps it has always been. What types of technologies will be foundational to our business? The disruption of Blockchain to data and databases parallels the disruption of the Internet to communications and networks. From Wall Street to Fintech Accelerators, incredible investments at Blockchain have led to incredible speed of innovation. This disruption is now ready to penetrate enterprise IT.

Which authors of this paper are endorsers? Disable MathJax (What is MathJax?)

Using Facebook Research's new fastText library in supervised mode, I trained a hybrid recommender system, to recommend articles to users, given as training data both the text in the articles and the user/article interaction matrix. The labels attached to a document were both its id, and the ids of all users who viewed it. I've not finished testing it, but early signs are that it learns quite well. This post is a short progress report, perhaps of interest to others building hybrid recommenders. Recently, Facebook Research released the C source code for fastText, a software library for learning word embeddings using a shallow neural network.

Indeed, in-person communication and touch and feel of the products in stores have not been topped by any achievements of e-commerce yet. And just like AI was able to significantly enhance the online experience with sophisticated recommendations algorithms and chatbots, there is a place for it in the physical world as well. Robotics powered by advanced NLP capabilities could be the next step in retail that will bring online and offline shopping experience much closer. In fact, the natural language processing market's size is estimated to grow from 7.63 billion in 2016 to 16.07 billion by 2021, at a CAGR of over 16%.

With all the talk here at Microsoft's Ignite conference around the cloud, office productivity and collaboration (and all that goes with that), CEO Satya Nadella took the keynote stage yesterday afternoon to discuss what he sees as the next horizon of digital transformation: artificial intelligence. This, to be sure, is not about AI for the sake of AI. "We don't want to pursue AI to beat humans at games," he declared with a winking swipe at IBM's Watson supercomputer, which won a Jeopardy Tournament, and Google's AlphaGo, which defeated a world champion at the game of Go. "We want to empower people with the tools of AI so they can build their own solutions." Microsoft's approach to AI, Nadella went on, is about infusing every application with intelligence. He noted that prior to the invention of the printing press with moveable type, there were but 13,000 books in the world. A few years later, there were 12 million books.

Is the Internet of Things the world's most confusing tech trend? On the one hand, we're told it's going to be epic, and soon – all predictions are either in tens of billions (of connected devices) and trillions (of dollars of economic value to be created). On the other hand, the dominant feeling expressed by end users (including at this year's CES show, arguably the bellwether of the industry) is essentially "meh" – right now the IoT feels like an avalanche of new connected products, many of which seem to solve trivial, "first world" problems: expensive gadgets that resolutely fall in the "nice to have" category, rather than "must have". And, for all the talk about a mega tech trend, things seem to be moving at the speed of molasses, with little discernible progress year on year. Part of the problem is perhaps one of semantics. While gadgets are indeed part of the category (and quite often very large markets onto themselves), the Internet of Things (which we define as any "connected hardware" other than desktops, laptops and smartphones) is a much broader, and deeper, trend that cuts across both the consumer, enterprise and industrial spaces.

Artificial intelligence (AI) isn't new but the rise of mobile and cloud computing, combined with big data and cheap computing power, is driving a resurgence. Convergent technologies mean AI is finding new uses in financial services. AI will be used in "every single segment of financial services", predicts Christophe Chazot, group head of innovation, HSBC. "The software is getting more intelligent in a human sense, mimicking human reasoning." The technology can help wealth advisors, back office staff and operations, traders and corporate finance teams.

This reference is a part of a new series of DSC articles, offering selected tutorials on subjects such as deep learning, machine learning, data science, deep data science, artificial intelligence, Internet of Things, algorithms, and related topics. It is designed for the busy reader who does not have a lot of time digging into long lists of advanced publications.

But if, as they say in this electoral season, you're looking to "grow your base," exercising emotional intelligence -- expressing empathy, checking your emotions in a bid to avoid conflict, and investing in personal relationships -- is a strategy that beats narcissism over the long term. A new exploration of how we make friends and influence people rigorously measured the emergence of popularity in small groups -- first-year college students organized into 15 study groups of about 20 in Poland. In the first week of their assignment to a group and then again three months later, 170 of the freshmen named the person or people they most liked in their group. Upon recruitment into the study, each participant completed standard inventories assessing their narcissistic personality traits and gauging their emotional intelligence. The findings: When a group of strangers is thrown together, individuals who score high on narcissism enjoy an early surge of admiration, recognition and friendship among their peers.

We consider a supervised machine learning scenario where labels are provided by a heterogeneous set of teachers, some of which are mediocre, incompetent, or perhaps even malicious. We present an algorithm, built on the SVM framework, that explicitly attempts to cope with low-quality and malicious teachers by decreasing their influence on the learning process. Our algorithm does not receive any prior information on the teachers, nor does it resort to repeated labeling (where each example is labeled by multiple teachers). We provide a theoretical analysis of our algorithm and demonstrate its merits empirically. Finally, we present a second algorithm with promising empirical results but without a formal analysis.