Classification problems with a large number of labels arise in language modelling [Mikolov et al., 2013, Levy and Goldberg, 2014], recommender systems [Covington et al., 2016, Xu et al., 2016], and information retrieval [Agrawal et al., 2013, Prabhu and Varma, 2014]. Such large-output problems pose a core challenge: losses such as the softmax cross-entropy can be prohibitive to optimise, as they depend on the entire set of labels. Several works have thus devised negative sampling schemes for efficiently and effectively approximating such losses [Bengio and Senecal, 2008, Blanc and Rendle, 2018, Ruiz et al., 2018, Bamler and Mandt, 2020]. Broadly, negative sampling techniques sample a subset of "negative" labels, which are used to contrast against the observed "positive" labels. One further applies a suitable weighting on these "negatives", which ostensibly corrects the sampling bias introduced by the dependence on a random subset of labels. Intuitively, such bias assesses how closely a scheme approximates the unsampled loss on the full label set. This bias is well understood for sampled softmax schemes (see, e.g., Bengio and Senecal [2008]); surprisingly, however, far less is understood about other popular schemes, e.g., within-batch and uniform sampling (cf.

Virtual assistants such as Amazon's Alexa, Apple's Siri, Google Home, and Microsoft's Cortana, are becoming ubiquitous in our daily lives and successfully help users in various daily tasks, such as making phone calls or playing music. Yet, they still struggle with playful utterances, which are not meant to be interpreted literally. Examples include jokes or absurd requests or questions such as, "Are you afraid of the dark?", "Who let the dogs out?", or "Order a zillion gummy bears". Today, virtual assistants often return irrelevant answers to such utterances, except for hard-coded ones addressed by canned replies. To address the challenge of automatically detecting playful utterances, we first characterize the different types of playful human-virtual assistant interaction. We introduce a taxonomy of playful requests rooted in theories of humor and refined by analyzing real-world traffic from Alexa. We then focus on one node, personification, where users refer to the virtual assistant as a person ("What do you do for fun?"). Our conjecture is that understanding such utterances will improve user experience with virtual assistants. We conducted a Wizard-of-Oz user study and showed that endowing virtual assistant s with the ability to identify humorous opportunities indeed has the potential to increase user satisfaction. We hope this work will contribute to the understanding of the landscape of the problem and inspire novel ideas and techniques towards the vision of giving virtual assistants a sense of humor.

Click-through rate (CTR) prediction is a critical problem in web search, recommendation systems and online advertisement displaying. Learning good feature interactions is essential to reflect user's preferences to items. Many CTR prediction models based on deep learning have been proposed, but researchers usually only pay attention to whether state-of-the-art performance is achieved, and ignore whether the entire framework is reasonable. In this work, we use the discrete choice model in economics to redefine the CTR prediction problem, and propose a general neural network framework built on self-attention mechanism. It is found that most existing CTR prediction models align with our proposed general framework. We also examine the expressive power and model complexity of our proposed framework, along with potential extensions to some existing models. And finally we demonstrate and verify our insights through some experimental results on public datasets.

Project Connected Home over IP (Project CHIP) is now known as Matter. The Connectivity Standards Alliance, an organization made up of more than a hundred device manufacturers, including giants like Apple, Amazon, Google and Samsung, announced the rebranding on Tuesday. Those companies came together to work on CHIP in 2019 with the hopes of building out an open smart home standard that connects all their disparate devices together. At its most simplest, the promise of Matter is that you'll be able to buy a device and use it with the voice assistant of your choice and easily connect it to your existing home network. At launch, Matter will support Alexa, Google Assistant, Siri, as well as Ethernet, WiFi, Thread and Bluetooth LE. "The Matter mark will serve as a seal of approval, taking the guesswork out of the purchasing process and allowing businesses and consumers alike to choose from a wider array of brands to create secure and connected homes and buildings," the CSA says of the rebranding.

The voice of Alexa, the virtual assistant developed by Amazon, is provided by Nina Rolle, a Colorado-based voiceover artist, according to a new book. Amazon has never revealed who provides the default female voice that responds to commands and questions given to Alexa, but the author Brad Stone said he identified the voice as Rolle's after "canvasing the professional voiceover community" for his new book, Amazon Unbound: Jeff Bezos and the Invention of a Global Empire. Rolle, who is based in Boulder, has conducted voiceover work for clients including Honda, Jenny Craig and Chase bank. According to Stone's book, she was selected after Amazon spent months assessing various candidates, with the final choice signed off by Jeff Bezos, the company's founder. Stone writes that Rolle said she was unable to talk about the role when he contacted her in February.

Machine learning has advanced from the age of science fiction to a major component of modern enterprises, especially as businesses across almost all sectors use various machine learning technologies. As an example, the healthcare industry is utilizing machine learning business applications to achieve more accurate diagnoses and provide better treatment to their patients. Retailers also use machine learning to send the right goods and products to the right stores before it is out of stock. Medical researchers are also not left out when it comes to using machine learning as many introduce newer and more effective medicines with the help of this technology. Many use cases are emerging from all sectors as machine learning is being implemented in logistics, manufacturing, hospitality, travel and tourism, energy, and utilities.

Cold-start problems are enormous challenges in practical recommender systems. One promising solution for this problem is cross-domain recommendation (CDR) which leverages rich information from an auxiliary (source) domain to improve the performance of recommender system in the target domain. In these CDR approaches, the family of Embedding and Mapping methods for CDR (EMCDR) is very effective, which explicitly learn a mapping function from source embeddings to target embeddings with overlapping users. However, these approaches suffer from one serious problem: the mapping function is only learned on limited overlapping users, and the function would be biased to the limited overlapping users, which leads to unsatisfying generalization ability and degrades the performance on cold-start users in the target domain. With the advantage of meta learning which has good generalization ability to novel tasks, we propose a transfer-meta framework for CDR (TMCDR) which has a transfer stage and a meta stage. In the transfer (pre-training) stage, a source model and a target model are trained on source and target domains, respectively. In the meta stage, a task-oriented meta network is learned to implicitly transform the user embedding in the source domain to the target feature space. In addition, the TMCDR is a general framework that can be applied upon various base models, e.g., MF, BPR, CML. By utilizing data from Amazon and Douban, we conduct extensive experiments on 6 cross-domain tasks to demonstrate the superior performance and compatibility of TMCDR.

The Amazon Echo Show 10 automatically moves its display to face the user, even if it is performing a task that doesn't need user input, like showing a recipe on the screen. Get weekly insights into the ways companies optimize data, technology and design to drive success with their customers and employees. Proactive or not, features in smart-home devices need to address a real user need, not stack the product with unnecessary and potentially confusing tools, said Ashton Udall, senior product manager at Google. The company developed sensor technology to monitor sleep, for example, because its research showed that consumers frequently forget to use or charge the wearables often employed for sleep tracking, or find the devices uncomfortable, he said. Amazon and Google hope the experiences will help them compete for users and more fully integrate their devices into people's lives.

Martin Taylor is the Deputy CEO and Co-Founder of Content Guru. Siri and Alexa -- the robots we couldn't live without. Throughout the pandemic, these voice assistants have proven invaluable to many, as users turned towards Alexa and Google Assistant for entertainment, education and emotional help. In fact, according to one survey, 3 in 5 users believe that their voice assistant has helped them get through isolation, and 40% will continue to use their digital assistants more as a result of the pandemic. These smart assistants are so effective because they're driven by artificial intelligence (AI).

Conversational recommender systems (CRSs) have revolutionized the conventional recommendation paradigm by embracing dialogue agents to dynamically capture the fine-grained user preference. In a typical conversational recommendation scenario, a CRS firstly generates questions to let the user clarify her/his demands and then makes suitable recommendations. Hence, the ability to generate suitable clarifying questions is the key to timely tracing users' dynamic preferences and achieving successful recommendations. However, existing CRSs fall short in asking high-quality questions because: (1) system-generated responses heavily depends on the performance of the dialogue policy agent, which has to be trained with huge conversation corpus to cover all circumstances; and (2) current CRSs cannot fully utilize the learned latent user profiles for generating appropriate and personalized responses. To mitigate these issues, we propose the Knowledge-Based Question Generation System (KBQG), a novel framework for conversational recommendation. Distinct from previous conversational recommender systems, KBQG models a user's preference in a finer granularity by identifying the most relevant relations from a structured knowledge graph (KG). Conditioned on the varied importance of different relations, the generated clarifying questions could perform better in impelling users to provide more details on their preferences. Finially, accurate recommendations can be generated in fewer conversational turns. Furthermore, the proposed KBQG outperforms all baselines in our experiments on two real-world datasets.