When I completed my degree in neuroinformatics in 2005, the next career step on the same topic was to go into research or to do a PhD. The areas of machine learning, cognitive science, natural language processing, robotics and all the other parts that make artificial intelligence where academic disciplines that had not hit the broad markets yet. Most companies at that time were still struggling to find out how to use the Internet. Remember, we are talking about web 1.0 with its static web pages, wonderful blinking captions, gif-buttons and online guestbooks. Happy about achieving my master's degree, I did not feel like becoming a scientist or continuing my studies, so (by coincidence) I started doing SAP consultancy. I do not want to say bad things about R/3, but at the time, going from neural nets coding that we did at the University of Edinburgh to ABAP felt like taking a step (or three) backwards in time and technology.

Is your dating experience starting to feel a little complicated? OKCupid is launching two fresh new features that could improve your dating chances in this complex new reality. The biggest renovation will bring mobile users a whole new way of viewing their potential matches. Instead of just seeing a photo, you'll now get a DoubleTake. It's a new layout that offers multiple photos and profile highlights.

In 1958, the Monotones asked, "Who Wrote the Book of Love?". If we fast forward 58 years to the present day, would the'Algorithm of Love' be a more accurate title? It is certainly a valid question; such is the popularity of the modern online dating website. Users can forego the inevitable awkwardness of blind or speed dating altogether, having found out every last detail about their date before they ever have to meet them in person. Most dating websites utilize some form of collaborative filtering or a'Netflix-style' recommendation algorithm that matches people with their potential partners based on shared interests and hobbies or mutual likes and dislikes.

Imagine you are a designer in a brainstorm meeting in Airbnb's headquarter in San Francisco. You are surrounded by posh art, brewed coffee, and hardcore liberal designers. The question on the white board is "How to make people of color feel more welcomed on Airbnb?" Imagine you are a designer working on OKCupid (or any other dating app). You open the emails from your marketing researcher and data scientist at the same time. The marketing data tell you that user likes to see bigger pictures and make a snap judgment.

Valentine's Day is widely celebrated as the day of love and friendship. It is the time of the year when people rekindle their romance, and if you don't have an ongoing one, you can use the opportunity to try and make someone your Valentine. Here are some apps that will help you find your ideal match on Valentine's Day: Hater: Compatibility is not a one-way road. Sometimes, you match up based on what you like and other times, you might even find your special someone based on something you both dislike. Hater is an app which asks users about their feelings on a variety of different topics and then does something different from what most dating apps do -- it matches people on what they dislike. According to the app's website, the most hated topics in the U.S. include slow drivers, mosquitoes, bullies, and not surprisingly, President Donald J. Trump.

Artificial Intelligence (AI) is often used as a plot device for doom-and-gloom science-fiction stories set in the future. The reality, however, is that AI has been around for quite some time--and it's incredibly useful for businesses. In fact, this technology has been assisting us daily in ways we barely even consider. Whenever you say, "Hey Siri," shop on Amazon, play songs on Pandora, or search for a photo on Facebook, AI is quietly working behind the scenes to deliver what you need. Over the past decade, this technology has slowly integrated into our everyday lives, but progress is rapidly picking up speed.

"Microsoft Monday" takes a look back at the past week of news related to Microsoft. This week, "Microsoft Monday" includes details about the acquisition of "Wiki" by Agile Extensions, a leaked Project NEON screenshot hinting at Windows 10 design changes, a federal judge ruling in favor of Microsoft in regards to the Justice Department lawsuit, the new suggested reminders feature for Cortana, Visual Studio 2017 launching on March 7th, the Surface Hub "Try & Buy Program" and more! In the past, Microsoft has been notorious for making design changes that did not sit well with consumers like we have seen in Windows Vista and Windows 8. Tom Hounsell, the creator of a Windows build tracker website called BuildFeed, published this leaked screenshot of the Windows 10 Project NEON design: The most notable Project NEON change is the removal of the window borders. This new user interface change could cause confusion in terms of resizing and moving windows. For example, some users may be confused about where to place the mouse in order to perform those functions.

In many time-aware item recommender systems, modeling the accurate evolution of both user profiles and the contents of items over time is essential. However, most existing methods focus on learning users' dynamic interests, where the contents of items are assumed to be stable over time. They thus fail to capture the dynamic changes in the item's contents. In this paper, we present a novel method CDUE for time-aware item recommendation, which captures the evolution of both user's interests and item's contents information via topic dynamics. Specifically, we propose a dynamic sparse topic model to track the evolution of topics for changes in items' contents over time and adapt a vector autoregressive model to profile users' dynamic interests. The item's topics and user's interests and their evolutions are learned collaboratively and simultaneously into a unified learning framework. Experimental results on two real-world data sets demonstrate the quality and effectiveness of the proposed method and show that our method can be used to make better future recommendations.

With the increasing popularity of location-aware social media applications, Point-of-Interest (POI) recommendation has recently been extensively studied. However, most of the existing studies explore from the users' perspective, namely recommending POIs for users. In contrast, we consider a new research problem of predicting users who will visit a given POI in a given future period. The challenge of the problem lies in the difficulty to effectively learn POI sequential transition and user preference, and integrate them for prediction. In this work, we propose a new latent representation model POI2Vec that is able to incorporate the geographical influence, which has been shown to be very important in modeling user mobility behavior. Note that existing representation models fail to incorporate the geographical influence. We further propose a method to jointly model the user preference and POI sequential transition influence for predicting potential visitors for a given POI. We conduct experiments on 2 real-world datasets to demonstrate the superiority of our proposed approach over the state-of-the-art algorithms for both next POI prediction and future user prediction.

Motivated by many real world applications such as recommendations in online shopping or entertainment, we consider the problem of selecting sequences of items. In this paper we introduce a novel class of utility functions over sequences of items, strictly generalizing the commonly used class of submodular set functions. We encode the sequential dependencies between items by a directed graph underlying the utility function. Classical algorithms fail to achieve any constant factor approximation guarantees on the problem of selecting sequences of bounded length with maximum utility. We propose an efficient algorithm for this problem that comes with strong theoretical guarantees characterized by the structural properties of the underlying graph. We demonstrate the effectiveness of our algorithm in synthetic and real world experiments on a movie recommendation dataset.