Voice technologies are slowly figuring out how to make themselves indispensable in our lives. While they aren't there yet, they likely WILL get there, and when they do the way we work and live will likely change dramatically. In this keynote at the Nashville Voice Conference hosted by [Data Driven Design] I talk about the benefit of being an early adopter and leaning into innovation.

This week Jordan Erica Webber is joined by Alex Hern, as they look at the scandal that rocked the voice assistant world, and ask whether or not we can trust that voice assistants aren't eavesdropping on our most private moments

It's no great secret that Apple's voice assistant has plenty of room for improvement. The Cupertino company is aware of this -- in June, it debuted an improved neural text-to-speech model capable of delivering a more natural-sounding voice without the use of samples. And in a newly published research paper on the preprint server Arxiv.org, a team of Apple scientists describe an approach for selecting training data for Siri's domain classifier -- the component that chooses whether a person's command relates to, say, their calendar rather than their alarms -- that leads to a substantial error reduction with only a small percentage of examples. As the researchers explain, Siri processes speech to suss out the intended domain with a classifier called the Domain Chooser, which helps identify a given user's intent. Once an utterance is matched to one of the over 60 defined domains, a component called the Statistical Parser assigns a parse label to each part of the utterance, after which the domain and parse labels predicted by the Domain Chooser and Statistical Parser are mapped into an intent representation that kicks off the appropriate action.

There is a tremendous amount of data generated today -- so much that our normal databases cannot manage. It is estimated that by 2020, every person will be generating 1.7 megabytes of data in just a single second. If you think 1.7MBs are small then you might be thinking about data in terms of storage. But this is in terms of storage; in simple terms, a single character like A, B or 7 accounts for 1 byte, a document containing only 100 characters without any overhead such as symbols would use 100 bytes. One megabyte contains 1,000,000 bytes or one million characters. This means every second one person will be generating 1.7 million characters and subsequently 102 million characters every minute or 6.1 billion characters every hour.

--In this paper, the method UCB-RS, which resorts to recommendation system (RS) for enhancing the upper-confidence bound algorithm UCB, is presented. The proposed method is used for dealing with non-stationary and large-state spaces multi-armed bandit problems. The proposed method has been targeted to the problem of the product recommendation in the online advertising. Through extensive testing with RecoGym, an OpenAI Gym-based reinforcement learning environment for the product recommendation in online advertising, the proposed method outperforms the widespread reinforcement learning schemes such as null -Greedy, Upper Confidence (UCB1) and Exponential Weights for Exploration and Exploitation (EXP3). I NTRODUCTION Online advertising is becoming increasingly popular and is the main motivation for the development of almost free internet platforms such as search engines, social networks, recruitment sites, multimedia contents (e.g., videos, images, musics, ...) sharing, etc. From the point of view of the internet users, the product recommendation on online advertising can be genuinely useful if it meets the real immediate needs of users. Instead of spending a lot of time and effort searching for a huge number of thousands or even millions of choices, most internet users will be quite satisfied if recommendation systems propose exactly what they need. Finding a good recommendation system, therefore, continues to be the goal of many studies [1], [2]. Online and offline approaches for learning optimal recommendation policies can be found in the literature.

Although numerous instantiations [ He et al., 2017; Liang et al., 2018 ] of CF have been proposed in recent years, matrix factorization (MF) [ Mnih and Salakhut-dinov, 2007; Koren et al., 2009 ] remains the most popular one due to its simplicity and effectiveness, and has been used for large scale recommendations of news [ Das et al., 2007], movies [ Koren et al., 2009 ] and products [ Linden et al., 2003 ] . Recent studies extend the MF framework for item cold-start recommendation by incorporating content information of items. The majority of methods for item cold-start recommendation employ a latent space sharing model. For example, Saveski te al. [ 2014] and Barjasteh et al. [ 2016 ] propose to use MF as the prjection function for both interactions and item contents. LDA [ Wang and Blei, 2011 ], CNN [ Kim et al., 2016 ], DNN [ Ebesu and Fang, 2017 ], SDAE [ Wang et al., 2015; Ying et al., 2016 ] and mDA [ Li et al., 2015 ] are proposed to learn the latent vectors of items from their textual contents. V an den Oord et al. [ 2013] and Wang et al. [ 2014] propose to use CNN to learn the latent vectors of music from their audio signals. The Wasserstein distance, which originates from optimal transport theory [ Rubner et al., 1998; Levina and Bickel, 2001], is a distance metric on probabilistic space and able to leverage the information on feature space. It has been successfully applied to many applications, such as computer vision [ Arjovsky et al., 2017 ] and natural language processing Figure 2: An illustration of problem definition.

Context-aware recommender systems (CARSs) apply sensing and analysis of user context in order to provide personalized services. Adding context to a recommendation model is challenging, since the addition of context may increases both the dimensionality and sparsity of the model. Recent research has shown that modeling contextual information as a latent vector may address the sparsity and dimensionality challenges. We suggest a new latent modeling of sequential context by generating sequences of contextual information and reducing their contextual space to a compressed latent space.We train a long short-term memory (LSTM) encoder-decoder network on sequences of contextual information and extract sequential latent context from the hidden layer of the network in order to represent a compressed representation of sequential data. We propose new context-aware recommendation models that extend the neural collaborative filtering approach and learn nonlinear interactions between latent features of users, items, and contexts which take into account the sequential latent context representation as part of the recommendation process. We deployed our approach using two context-aware datasets with different context dimensions. Empirical analysis of our results validates that our proposed sequential latent context-aware model (SLCM), surpasses state of the art CARS models.

Collaborative filtering, a widely-used recommendation technique, predicts a user's preference by aggregating the ratings from similar users. As a result, these measures cannot fully utilize the rating information and are not suitable for real world sparse data. To solve these issues, we propose a novel user distance measure named Preference Mover's Distance (PMD) which makes full use of all ratings made by each user. Our proposed PMD can properly measure the distance between a pair of users even if they have no co-rated items. We show that this measure can be cast as an instance of the Earth Mover's Distance, a well-studied transportation problem for which several highly efficient solvers have been developed. Experimental results show that PMD can help achieve superior recommendation accuracy than state-of-the-art methods, especially when training data is very sparse.

Traditional recommendation systems produce static rather than interactive recommendations invariant to a user's specific requests, clarifications, or current mood, and can suffer from the cold-start problem if their tastes are unknown. These issues can be alleviated by treating recommendation as an interactive dialogue task instead, where an expert recommender can sequentially ask about someone's preferences, react to their requests, and recommend more appropriate items. In this work, we collect a goal-driven recommendation dialogue dataset (GoRecDial), which consists of 9,125 dialogue games and 81,260 conversation turns between pairs of human workers recommending movies to each other. The task is specifically designed as a cooperative game between two players working towards a quantifiable common goal. We leverage the dataset to develop an end-to-end dialogue system that can simultaneously converse and recommend. Models are first trained to imitate the behavior of human players without considering the task goal itself (supervised training). We then finetune our models on simulated bot-bot conversations between two paired pre-trained models (bot-play), in order to achieve the dialogue goal. Our experiments show that models finetuned with bot-play learn improved dialogue strategies, reach the dialogue goal more often when paired with a human, and are rated as more consistent by humans compared to models trained without bot-play. The dataset and code are publicly available through the ParlAI framework.

The Internet of Things, or IoT, is a growing infrastructure of internet-enabled objects ranging from vacuums to light bulbs, all aimed at increasing control, automation and even data collection. IoT can be a huge benefit for a business office when used appropriately. Many offices are already used to internet-connected printers, but a new generation of smart alternatives is hitting the market that allow more than network printing. They monitor their paper and ink and can warn a support person when they're getting low. They can also connect to inventory systems to know how much spare ink or paper they have on hand and can even make orders for more without human involvement.