SHOPPING – Contains affiliated content. Products featured in this Mail Best article are selected by our shopping writers. If you make a purchase using links on this page, Dailymail.co.uk will earn an affiliate commission. In the world of smart speakers, the Echo Dot has earned its place as a fan favourite, and now you can get your hands on the latest and greatest model at an incredible 60 per cent off thanks to this unmissable Prime Day deal. The Echo Dot (5th generation, 2022 release) is now available for just £21.99, down from its original price of £54.99 - this Prime Day deal is too good to pass up.

Like the oozing black river that gave it its name, Amazon Prime Day flows through another year. It's spreading across the internet delta too--it's now two days long and last year there were two separate Prime Day events. This year's, ahem, prime Prime Day sales event starts now and runs through the end of July 12. We've spent hours combing through thousands of lists to find the absolute best Prime Day deals on WIRED-tested gear, from Alexa-enabled speakers and robot vacuums to laptops and tablets. Note: Many of these deals may require a membership, including Amazon Prime, Target Circle, or Walmart Plus. Amazon Prime has a free 30-day trial if you don't already have it (set a reminder to cancel it before it automatically renews). Target Circle is free to sign up, but you'll need to save the offer to your account from the main offer page or on the actual buy page for a particular item to see the deal price at checkout. Many of Walmart's deals are in early access for subscribers but will be available for everyone later in the day. If you buy something using links in our stories, we may earn a commission. This helps support our journalism. Our favorite speaker for the bedroom, the Echo Dot makes a great alarm clock. It'll wake you up to your favorite playlists and read you the weather if you like. The 4th gen Echo is our favorite Alexa speaker.

Machine learning is traditionally studied at the model level: researchers measure and improve the accuracy, robustness, bias, efficiency, and other dimensions of specific models. In practice, the societal impact of machine learning is determined by the surrounding context of machine learning deployments. To capture this, we introduce ecosystem-level analysis: rather than analyzing a single model, we consider the collection of models that are deployed in a given context. For example, ecosystem-level analysis in hiring recognizes that a job candidate's outcomes are not only determined by a single hiring algorithm or firm but instead by the collective decisions of all the firms they applied to. Across three modalities (text, images, speech) and 11 datasets, we establish a clear trend: deployed machine learning is prone to systemic failure, meaning some users are exclusively misclassified by all models available. Even when individual models improve at the population level over time, we find these improvements rarely reduce the prevalence of systemic failure. Instead, the benefits of these improvements predominantly accrue to individuals who are already correctly classified by other models. In light of these trends, we consider medical imaging for dermatology where the costs of systemic failure are especially high. While traditional analyses reveal racial performance disparities for both models and humans, ecosystem-level analysis reveals new forms of racial disparity in model predictions that do not present in human predictions. These examples demonstrate ecosystem-level analysis has unique strengths for characterizing the societal impact of machine learning.

We consider the problem of latent bandits with cluster structure where there are multiple users, each with an associated multi-armed bandit problem. These users are grouped into \emph{latent} clusters such that the mean reward vectors of users within the same cluster are identical. At each round, a user, selected uniformly at random, pulls an arm and observes a corresponding noisy reward. The goal of the users is to maximize their cumulative rewards. This problem is central to practical recommendation systems and has received wide attention of late \cite{gentile2014online, maillard2014latent}. Now, if each user acts independently, then they would have to explore each arm independently and a regret of $\Omega(\sqrt{\mathsf{MNT}})$ is unavoidable, where $\mathsf{M}, \mathsf{N}$ are the number of arms and users, respectively. Instead, we propose LATTICE (Latent bAndiTs via maTrIx ComplEtion) which allows exploitation of the latent cluster structure to provide the minimax optimal regret of $\widetilde{O}(\sqrt{(\mathsf{M}+\mathsf{N})\mathsf{T}})$, when the number of clusters is $\widetilde{O}(1)$. This is the first algorithm to guarantee such strong regret bound. LATTICE is based on a careful exploitation of arm information within a cluster while simultaneously clustering users. Furthermore, it is computationally efficient and requires only $O(\log{\mathsf{T}})$ calls to an offline matrix completion oracle across all $\mathsf{T}$ rounds.

Personalized recommendations have a growing importance in direct marketing, which motivates research to enhance customer experiences by knowledge graph (KG) applications. For example, in financial services, companies may benefit from providing relevant financial articles to their customers to cultivate relationships, foster client engagement and promote informed financial decisions. While several approaches center on KG-based recommender systems for improved content, in this study we focus on interpretable KG-based recommender systems for decision making.To this end, we present two knowledge graph-based approaches for personalized article recommendations for a set of customers of a large multinational financial services company. The first approach employs Reinforcement Learning and the second approach uses the XGBoost algorithm for recommending articles to the customers. Both approaches make use of a KG generated from both structured (tabular data) and unstructured data (a large body of text data).Using the Reinforcement Learning-based recommender system we could leverage the graph traversal path leading to the recommendation as a way to generate interpretations (Path Directed Reasoning (PDR)). In the XGBoost-based approach, one can also provide explainable results using post-hoc methods such as SHAP (SHapley Additive exPlanations) and ELI5 (Explain Like I am Five).Importantly, our approach offers explainable results, promoting better decision-making. This study underscores the potential of combining advanced machine learning techniques with KG-driven insights to bolster experience in customer relationship management.

In graph representation learning, it is important that the complex geometric structure of the input graph, e.g. hidden relations among nodes, is well captured in embedding space. However, standard Euclidean embedding spaces have a limited capacity in representing graphs of varying structures. A promising candidate for the faithful embedding of data with varying structure is product manifolds of component spaces of different geometries (spherical, hyperbolic, or euclidean). In this paper, we take a closer look at the structure of product manifold embedding spaces and argue that each component space in a product contributes differently to expressing structures in the input graph, hence should be weighted accordingly. This is different from previous works which consider the roles of different components equally. We then propose WEIGHTED-PM, a data-driven method for learning embedding of heterogeneous graphs in weighted product manifolds. Our method utilizes the topological information of the input graph to automatically determine the weight of each component in product spaces. Extensive experiments on synthetic and real-world graph datasets demonstrate that WEIGHTED-PM is capable of learning better graph representations with lower geometric distortion from input data, and performs better on multiple downstream tasks, such as word similarity learning, top-$k$ recommendation, and knowledge graph embedding.

Your article (Never underestimate a droid: robots gather at AI for Good summit in Geneva, 6 July) begins by listing four of the robot delegates that are attending the AI for Good summit – all four are "feminised robots" – and I remembered the thought I had when I saw Ai-Da perform poetry at the Ashmolean in Oxford in 2021: why does a robot need boobs? Robotics and AI are fields undoubtedly occupied primarily by men and yet many robots, and AI assistants (think Siri, Alexa and so on) often take on a "feminised" form. Perhaps we are more comfortable telling a feminised voice to do things for us. Ai-Da and Desdemona, robot artist and musician respectively, have both been performing and receiving accolades, and both were created by men. Yet real-life female artists and musicians still struggle to get equal respect and representation in their fields. It is no great revelation to say that all AI will reflect society – including its misogyny – but entrenching gender roles and sexualisation of the female form should be highlighted before the advancement of robotics deepens them even further.

In May, Sydney Leathers confessed to her tens of thousands of Twitter followers that she was smitten. Where'd she meet the guy? Not on a dating app, or through friends, but in the last place she ever expected to find a real connection: her mentions. "Still can't believe I fell in love with one of my reply guys. Apparently, things had progressed since December, when she last posted about him: "I had sex with someone who started as my reply guy and I hope this doesn't inspire confidence in the rest of you because frankly your replies are not that good," she wrote. Leathers is a writer, adult performer, and startup employee whose name you may recognize from her part in the Anthony Weiner sexting scandal--this wasn't exactly her first brush with online flirtation. But it was her first time falling for a reply guy, or someone who was, effectively, a fan. The term "reply guy" emerged on Twitter about five years ago to describe the behavior of a certain subset of people, usually with very few social media followers of their own, who stake out space in the mentions of prominent users. They can be counted on to reply promptly and frequently to the tweets of whomever they've chosen as their object of devotion, and they often seek attention by nitpicking, mansplaining, joke one-upping, and harassing them. Because of this, reply guys--who can also be girls, or people of any gender--are generally understood to be pathetic creatures, without a chance in hell of getting said person to like their replies, much less return their affections. So the revelation that this gambit actually worked for someone is … pretty noteworthy. Reply guy success stories may be happening more than we realize. Abby, a 25-year-old in Brooklyn who runs a meme page on Instagram with several thousand followers, told me that she got frisky with one of her reply guys last year. "I'm not the only person that I know that has hooked up with reply guys," she said. "It's not as uncommon as you might think." Now, Leathers' Twitter feed is a monument to her relationship, by turns adorable and lewd. "This definitely caught me by surprise," she told me. "But it's been the best, happiest relationship I've had." To attain this goal, a reply guy's first challenge is to stand out from the crowd. The meme account Abby is the admin for is about politics, so she likes when a guy can show not just that he's hot, but that they share a political sensibility. "I have to be attracted to them," she said. "And they have to have some sort of compelling thing to say." "I feel like I've never more than mildly acknowledged a reply guy before now," she said. "I generally don't even follow them back." But when her now-boyfriend started responding to her tweets last year after discovering her through a winding path that involved the singer of the band Eve 6, she took notice. "I'd seen him reply to my stuff a few times.

This paper presents a solution to the challenges faced by contrastive learning in sequential recommendation systems. In particular, it addresses the issue of false negative, which limits the effectiveness of recommendation algorithms. By introducing an advanced approach to contrastive learning, the proposed method improves the quality of item embeddings and mitigates the problem of falsely categorizing similar instances as dissimilar. Experimental results demonstrate performance enhancements compared to existing systems. The flexibility and applicability of the proposed approach across various recommendation scenarios further highlight its value in enhancing sequential recommendation systems.

Our paper investigates the use of discourse embedding techniques to develop a community recommendation system that focuses on mental health support groups on social media. Social media platforms provide a means for users to anonymously connect with communities that cater to their specific interests. However, with the vast number of online communities available, users may face difficulties in identifying relevant groups to address their mental health concerns. To address this challenge, we explore the integration of discourse information from various subreddit communities using embedding techniques to develop an effective recommendation system. Our approach involves the use of content-based and collaborative filtering techniques to enhance the performance of the recommendation system. Our findings indicate that the proposed approach outperforms the use of each technique separately and provides interpretability in the recommendation process.