The phrase "artificial intelligence (AI)" carries as much heft as any in our marketing lexicon in 2019. Innovation surrounding AI and machine learning has flourished over the past decade. There is seemingly no limit to the opportunities and potential for AI to impact and improve the customer experience. Despite how far we have come in this area, especially as it pertains to marketing and advertising, we have still only scratched the surface. Consumers can only adopt so much change at one time. Move too slowly and your competition will begin to capitalize on the wealth of data available.

This paper introduces Seeker, a system that allows users to interactively refine search rankings in real time, through feedback in the form of likes and dislikes. When searching online, users may not know how to accurately describe their product of choice in words. An alternative approach is to search an embedding space, allowing the user to query using a representation of the item (like a tune for a song, or a picture for an object). However, this approach requires the user to possess an example representation of their desired item. Additionally, most current search systems do not allow the user to dynamically adapt the results with further feedback. On the other hand, users often have a mental picture of the desired item and are able to answer ordinal questions of the form: "Is this item similar to what you have in mind?" With this assumption, our algorithm allows for users to provide sequential feedback on search results to adapt the search feed. We show that our proposed approach works well both qualitatively and quantitatively. Unlike most previous representation-based search systems, we can quantify the quality of our algorithm by evaluating humans-in-the-loop experiments.

Not all voice assistants can handle the same requests. We put Siri, Alexa and Google to the test. Personal digital assistants – the voice-based interfaces that started with Apple's Siri – were supposed to enable entirely new and significantly more intuitive ways of interacting with our devices, but they ended up being so frustrating that many people gave up on them. Thankfully, we're starting to see some significant advances in phone-based digital assistants, and recent developments from both Microsoft's Build and Google's I/O (and likely from Apple's upcoming WWDC) developer conferences all highlight the important progress that has been made. These conferences, primarily intended for software developers to learn about new advances in the host company's technology platforms, also serve as a great guidepost for consumers to understand how technology is evolving.

People care about their privacy, but not enough to quit buying gadgets that expose their personal data, says a new study on consumer habits. Consumers International and the Internet Society surveyed thousands of people across North America, Europe and Asia to better understand the relationship between consumers and'smart' devices -- a term they defined as'everyday device and products that can connect to the internet.' The study did not include phones and mobile apps, which present a vast and more complex array of privacy issues. What they found was that many respondents using products like Google Home or Amazon Echo, fitness wearables, gaming consoles and internet-connected home appliances shared concern over how those devices harvest and share their personal data. Consumers think that data collection of their smart devices is'creepy' according to a new study of consumers across the world.

Match is becoming the first major dating app to provide its premium users with personally-tailored advice through a free human coach. The company announced today that it is beginning to roll out a new service called AskMatch which allows its paid users to chat on the phone with one of the company's dating hired'experts.' According to a report from TechCrunch, Match members can pick their coach's brains on a variety of topics that include how to set up a good dating profile, getting over a break up, or more general advice on dating. In multiple phone interviews, Match CEO, Hesam Hosseini said that the service will help to push the online dating platform, which has been in existence since 1995, into the future. 'Match's mission has always been around relationships and bringing people together.

In a gallery in downtown Manhattan, people are huddling around four laptops, taking turns to control the apartments of 14 complete strangers. They watch via live video feeds, and respond whenever the residents ask "Someone" to help them. They switch the lights on and off, boil the kettle, put some music on – whatever they can do to oblige. The project, called Someone, is the latest in a series exploring our ever more complicated relationship with technology. It's by the American artist Lauren McCarthy and is a sort of outsourcing of Lauren, an earlier work in which she acted as a real-life Alexa, remotely watching over a home 24 hours a day, responding to its occupants' questions and needs like a flesh and blood version of Amazon's voice-operated virtual assistant. Lauren, a video work, features in AI: More Than Human, which opens this week at the Barbican in London as part of its Life Rewired season, an investigation into what it means to be human in the digital era.

In a variety of online settings involving interaction with end-users it is critical for the systems to adapt to changes in user preference. User preferences on items tend to change over time due to a variety of factors such as change in context, the task being performed, or other short-term or long-term external factors. Recommender systems, in particular need to be able to capture these dynamics in user preferences in order to remain tuned to the most current interests of users. In this work we present a recommendation framework which takes into account the dynamics of user preferences. We propose an approach based on Hidden Markov Models (HMM) to identify change-points in the sequence of user interactions which reflect significant changes in preference according to the sequential behavior of all the users in the data. The proposed framework leverages the identified change points to generate recommendations using a sequence-aware non-negative matrix factorization model. We empirically demonstrate the effectiveness of the HMM-based change detection method as compared to standard baseline methods. Additionally, we evaluate the performance of the proposed recommendation method and show that it compares favorably to state-of-the-art sequence-aware recommendation models.

For example, while a recommendation of their behavior such as browsing trails, bookmarks ratings, source based on co-authorship links ranks attendees or created social links. It enables modern recommender systems by its social similarity with the target user, the recommendation to use multiple sources of information about user interests case might require to find attendees who are interested and preferences to deliver better recommendations. This in similar topics while being most likely unknown is most frequently done using parallel hybrid recommendation to the target user (i.e., having the weakest social similarity).

The accuracy of Top-N recommendation task is challenged in the systems with mainly implicit user feedback considered. Adversarial training has presented successful results in identifying real data distributions in various domains (e.g. image processing). Nonetheless, adversarial training applied to recommendation is still challenged especially by interpretation of negative implicit feedback causing it to converge slowly as well as affecting its convergence stability. This is often attributed to high sparsity of the implicit feedback and discrete values characteristic from items recommendation. To face these challenges, we propose a novel model named convolutional adversarial latent factor model (CALF), which uses adversarial training in generative and discriminative models for implicit feedback recommendations. We assume that users prefer observed items over generated items and then apply pairwise product to model the user-item interactions. Additionally, the latent features become input data of our convolutional neural network (CNN) to learn correlations among embedding dimensions. Finally, Rao-Blackwellized sampling is adopted to deal with the discrete values optimizing CALF and stabilizing the training step. We conducted extensive experiments on three different benchmark datasets, where our proposed model demonstrates its efficiency for item recommendation.

Many recommender systems suffer from popularity bias: popular items are recommended frequently while less popular, niche products, are recommended rarely or not at all. However, recommending the ignored products in the ``long tail'' is critical for businesses as they are less likely to be discovered. In this paper, we introduce a personalized diversification re-ranking approach to increase the representation of less popular items in recommendations while maintaining acceptable recommendation accuracy. Our approach is a post-processing step that can be applied to the output of any recommender system. We show that our approach is capable of managing popularity bias more effectively, compared with an existing method based on regularization. We also examine both new and existing metrics to measure the coverage of long-tail items in the recommendation.