Don't expect to find a Tinder date in the metaverse any time soon. The Verge reports Match Group chief Bernard Kim has asked Tinder's Hyperconnect unit (acquired in 2021) to scale back its metaverse dating plans. In his shareholder letter, Kim said "uncertainty" about success with virtual worlds required that the team "not invest heavily" in the metaverse. Match further blamed the Hyperconnect purchase for a $10 million operating loss in the latest quarter where it made a $210 million operating profit in the same period a year earlier. The company is also taking "a step back" on plans to introduce its in-app Tinder Coins following questionable test results, Kim said.

The chief executive of Tinder has left the dating app after less than a year after the market value of its parent company plunged by more than a fifth following reporting disappointing results. The departure of Renate Nyborg was one of a number of management changes announced by the $20bn Match Group, which owns dating brands including Hinge, Tinder and Match.com. Its share price plunged by more than 20% on Tuesday after missing Wall Street second-quarter expectations and issuing weaker-than-expected guidance. "Today we are announcing the departure of Tinder chief executive Renate Nyborg, and I have made some changes to the management team and structure that I am confident will help deliver Tinder's full potential," said Bernard Kim, the chief executive of Match Group, in a letter to shareholders. "We have not been able to realise the monetisation successes that we typically deliver. Tinder's current revenue growth expectations for the second half of the year are below our original expectations as a result of disappointing execution on several optimisations and new product initiatives."

"Today we're announcing the departure of Tinder CEO Renate Nyborg, and I have made some changes to the management team and structure that I am confident will help deliver Tinder's full potential," Match Group chief executive Bernard Kim said in a letter to shareholders.

Recently, graph neural networks have shown the superiority of modeling the complex topological structures in heterogeneous network-based recommender systems. Due to the diverse interactions among nodes and abundant semantics emerging from diverse types of nodes and edges, there is a bursting research interest in learning expressive node representations in multiplex heterogeneous networks. One of the most important tasks in recommender systems is to predict the potential connection between two nodes under a specific edge type (i.e., relationship). Although existing studies utilize explicit metapaths to aggregate neighbors, practically they only consider intra-relationship metapaths and thus fail to leverage the potential uplift by inter-relationship information. Moreover, it is not always straightforward to exploit inter-relationship metapaths comprehensively under diverse relationships, especially with the increasing number of node and edge types. In addition, contributions of different relationships between two nodes are difficult to measure. To address the challenges, we propose HybridGNN, an end-to-end GNN model with hybrid aggregation flows and hierarchical attentions to fully utilize the heterogeneity in the multiplex scenarios. Specifically, HybridGNN applies a randomized inter-relationship exploration module to exploit the multiplexity property among different relationships. Then, our model leverages hybrid aggregation flows under intra-relationship metapaths and randomized exploration to learn the rich semantics. To explore the importance of different aggregation flow and take advantage of the multiplexity property, we bring forward a novel hierarchical attention module which leverages both metapath-level attention and relationship-level attention. Extensive experimental results suggest that HybridGNN achieves the best performance compared to several state-of-the-art baselines.

Amazon Prime Day is now over - for July, anyway. According to leaked notices there may be another Prime Day deals extravaganza in Q4 later this year, as we reported earlier, but for now there are still plenty of discounts on everything from TVs to toys, and we've rounded up the best ones here. Jump to section: 1. Today's best deals 2. Amazon device deals 3. Back to school deals 4. TV deals 5. Laptop deals 6. Every choice on our list has been picked out by our expert editorial team here at TechRadar and we stand by our recommendations. We're well aware that everyone is feeling the pressure this year with inflation, gas prices, and many other household essentials becoming more expensive. We've made sure we've included bargains on a range of useful products that won't break the bank, as well as our old favorites like premium TVs. Where applicable, we'll always tell you if a Prime Day deal is at its lowest ever price or not too. Now that the main Amazon Prime Day 2022 sale is over, you no longer need to be a member to take advantage of today's offers. If you no longer want or need to keep your subscription going, make sure you find out how to cancel Amazon Prime while it's ongoing so you aren't charged for another month. With all that covered, let's head on into TechRadar's official list of the best Amazon Prime Day deals of 2022 that you can still buy. Apple Watch SE (40mm, GPS): $279 $219 at Amazon (opens in new tab) Save $70 - On a budget? We highly recommend the Apple Watch SE as the best smartwatch (opens in new tab) as the affordable iOS option. While it doesn't have as big of a screen or some of the advanced ECG and blood oxygen monitoring as the Apple Watch 7, the SE is still a great choice. It still has all those fitness, health, and lifestyle apps that the Apple Watches are famed for. Waterpik Aquarius dental flosser: $99.99 $54.99 at Amazon (opens in new tab) Save $45 - Amazon is currently offering this Waterpik Aquarius dental flosser for just over half price. It was $10 cheaper on Prime Day itself, but a 40% discount still isn't bad. It comes with 10 customization settings, a built-in timer to help you track flossing time and even a massage mode for gum stimulation.

Artificial Intelligence (AI) and machine learning are now accessible to WordPress users through AI-powered plugins. Currently, 77% of consumers use AI technology services or products, whether they are aware of it or not. It includes intelligent WordPress plugins. AI applications help boost businesses running on WordPress, the most popular Content Management System (CMS) platform powering over 35% of all websites to date. Initially a blogging platform, WordPress has become an extensive CMS, compared to Drupal or Joomla, and other similar platforms.

The past few years have seen a massive boom in smartphone technology and applications, mainly due to the rapid evolution in artificial intelligence and related technologies. AI has managed to completely transform the user interface and experience for even the simplest mobile apps that we use on a daily basis. In today's world, half of our daily tasks are automated and performed by mobile apps that are powered by AI, and it is truly a magical time to be alive. Speaking of artificial intelligence, it is impertinent to mention that the ever-growing phenomenon continues to morph mobile apps into something entirely different. From physical learning to AI-based learning, from physical piano lessons to online lessons for piano – everything is dominated by AI.

Recently, a former Google researcher claimed that some algorithms used by the company reached sentient capabilities well above their initial design. Phillipa Louvois rules that Data, the Enterprise's android, is not the property of Starfleet, arguing: "We have all been dancing around the basic issue: does Data have a soul? I don't know that he has. I don't know that I have. But I have got to give him the freedom to explore that question himself."

Next Point-of-Interest (POI) recommendation has become an indispensable functionality in Location-based Social Networks (LBSNs) due to its effectiveness in helping people decide the next POI to visit. However, accurate recommendation requires a vast amount of historical check-in data, thus threatening user privacy as the location-sensitive data needs to be handled by cloud servers. Although there have been several on-device frameworks for privacy-preserving POI recommendations, they are still resource-intensive when it comes to storage and computation, and show limited robustness to the high sparsity of user-POI interactions. On this basis, we propose a novel decentralized collaborative learning framework for POI recommendation (DCLR), which allows users to train their personalized models locally in a collaborative manner. DCLR significantly reduces the local models' dependence on the cloud for training, and can be used to expand arbitrary centralized recommendation models. To counteract the sparsity of on-device user data when learning each local model, we design two self-supervision signals to pretrain the POI representations on the server with geographical and categorical correlations of POIs. To facilitate collaborative learning, we innovatively propose to incorporate knowledge from either geographically or semantically similar users into each local model with attentive aggregation and mutual information maximization. The collaborative learning process makes use of communications between devices while requiring only minor engagement from the central server for identifying user groups, and is compatible with common privacy preservation mechanisms like differential privacy. We evaluate DCLR with two real-world datasets, where the results show that DCLR outperforms state-of-the-art on-device frameworks and yields competitive results compared with centralized counterparts.

Modeling the evolution of user preference is essential in recommender systems. Recently, dynamic graph-based methods have been studied and achieved SOTA for recommendation, majority of which focus on user's stable long-term preference. However, in real-world scenario, user's short-term preference evolves over time dynamically. Although there exists sequential methods that attempt to capture it, how to model the evolution of short-term preference with dynamic graph-based methods has not been well-addressed yet. In particular: 1) existing methods do not explicitly encode and capture the evolution of short-term preference as sequential methods do; 2) simply using last few interactions is not enough for modeling the changing trend. In this paper, we propose Long Short-Term Preference Modeling for Continuous-Time Sequential Recommendation (LSTSR) to capture the evolution of short-term preference under dynamic graph. Specifically, we explicitly encode short-term preference and optimize it via memory mechanism, which has three key operations: Message, Aggregate and Update. Our memory mechanism can not only store one-hop information, but also trigger with new interactions online. Extensive experiments conducted on five public datasets show that LSTSR consistently outperforms many state-of-the-art recommendation methods across various lines.