Content-based collaborative filtering (CCF) predicts user-item interactions based on both users' interaction history and items' content information. Recently, pre-trained language models (PLM) have been used to extract high-quality item encodings for CCF. However, it is resource-intensive to train a PLM-based CCF model in an end-to-end (E2E) manner, since optimization involves back-propagating through every content encoding within a given user interaction sequence. To tackle this issue, we propose GRAM (GRadient Accumulation for Multi-modality in CCF), which exploits the fact that a given item often appears multiple times within a batch of interaction histories. Specifically, Single-step GRAM aggregates each item encoding's gradients for back-propagation, with theoretic equivalence to the standard E2E training. As an extension of Single-step GRAM, we propose Multi-step GRAM, which increases the gradient update latency, achieving a further speedup with drastically less GPU memory. GRAM significantly improves training efficiency (up to 146x) on five datasets from two task domains of Knowledge Tracing and News Recommendation. Our code is available at https://github.com/yoonseok312/GRAM.

The AI market is expected to be worth $36.8 billion by 2025, and as this autonomous branch of digital technology continues to evolve, we expect to see a host of incredible developments to emerge in the not-so-distant future--especially in this emerging age of the metaverse. While AI has been proven to enhance our daily consumer lives--with the adoption of autonomous vehicle technology, fitness devices, voice assistants, and smartphones--it has also served to improve the digital marketing efforts of businesses across many industries. Not only is AI proven to boost a company's productivity by up to 40%, but if leveraged strategically, autonomous technology can help brands to: Indeed, if utilized tactically and creatively, AI technology has the potential to help you connect with your audience on a deeper level than ever before, resulting in ongoing business growth. To put this notion into perspective, here we explore some quite different but equally inspiring uses of AI in digital marketing. The first of our inspiring AI in digital marketing examples comes from Sephora. Understanding the power of AI early on, thriving cosmetics brand Sephora has been using chatbot technology to offer its audience the most personalized, informational customer experience possible.

The smart home ecosystem is getting a little more integrated: Google just updated its Amazon Alexa Skill to work on its latest Nest cameras. So if you already own a variety of Nest and Alexa devices, they'll work together more seamlessly moving forward. Now, you can stream live feeds from your Nest cameras, doorbells and other devices to anything from your Amazon Fire TV to any Echo device. Amazon made a similar gesture to open up its own smart home ecosystem last month, when it announced that its doorbells and security cameras would work with Google Nest, Ring, Abode and other third-party devices. Amazon's Ring doorbell already works with Google Home and Apple Homekit.

Google said a year ago that Assistant would offer to change compromised passwords, and now the company is delivering on that promise in earnest. Android Police notes that Assistant warnings are now rolling out more broadly to Chrome users, including those on Android. Sign in to a site with a compromised password and you'll both get the familiar "change your password" alert and, on some sites, an option to have Assistant automatically make that change. You can take control at any point, but this will be helpful if you'd rather not spend time devising a secure replacement sign-in. The helper uses the AI from Google's web version of Duplex to navigate websites and apply password changes.

Products featured in this Mail Best article are independently selected by our shopping writers. If you make a purchase using links on this page, MailOnline may earn an affiliate commission. Amazon has launched a massive sale with unmissable discounts on their bestselling Echo devices. Ahead of Amazon Prime Day, rumoured to be taking place July 18 for just 24 hours, shoppers can score a new Echo Dot smart speaker, Echo Show 8, Echo Flex and more and save up to 60 per cent off. Popular offers include the number one bestselling Echo Dot (4th Gen) on sale for £27.99 (was £49.99), and Amazon Echo Dot (3rd Gen) is now 35 per cent off, so you can score the hockey puck-shaped smart speaker for just £25.99 (was £39.99).

It seems Sonos is gearing up to roll out its own long-rumored voice assistant in the coming weeks. Sonos Voice is said to offer voice control for music playback on many of the company's devices, offering owners another option if they'd rather not use Amazon Alexa and Google Assistant. Sonos will first roll out Sonos Voice in the US on June 1st as part of a software update, according to The Verge. The feature should arrive in other countries later. A rumored $250 soundbar called Sonos Ray will likely be among those.

You may also have heard machine learning and AI used interchangeably. AI includes machine learning, but machine learning doesn't fully define AI. Machine learning and AI both have strong engineering components. You find AI and machine learning used in a great many applications today. Artificial Intelligence (AI) is a huge topic today, and it's getting bigger all the time thanks to the success of technologies such as Siri.

In my blog "The Importance of Determining Intent", I discussed the importance of determining user intent to create an "intelligent" user or stakeholder experience. Analytics-centric organizations specialize in determining and codifying a user's intent in order to provide a more engaging, relevant, hyper-personalized experience (Figure 1). Figure 1: Using "Intent Determination" to Create an Intelligent Customer Experience To create an "intelligent" user experience requires leveraging AI/ML to analyze a deep history of the user's interactions to determine the user's intentions, and then coupling those intentions with current trends, patterns, and relationships to match those intentions with a deep understanding of the available content to recommend the most relevant action. We reviewed how digital marketing companies, such as those featured in Figure 1, determine user intent. These companies accumulate a deep history of each individual user's interactions including what sites or content they visited or viewed, how long they spent with each site or piece of content, what they clicked on, what they did not click on, and their contextual search requests. They analyze the user's interaction history, and match that with current trends and behaviors of similar cohorts, to determine and codify (think propensity scores) the user's intentions (areas of interest) that drives real-time recommendation decisions.

But something clearly went haywire when the primary user of the speaker subscribes to YouTube Music and a second user turns on Voice Match. Things make more sense with Amazon Alexa and Apple Siri, which also have voice capabilities. I checked in with both companies, and neither cuts off other members of a household with voice matches from using a shared streaming music account.

A huge reason behind Sonos' success is the company's adaptability. Whenever a new audio streaming service comes along -- and as the industry evolves and pushes new formats like lossless and Dolby Atmos spatial audio -- Sonos is always right there trying to support what's new. For voice assistants, Sonos offers both Amazon Alexa and Google Assistant on its smart speakers. And the company also widely supports Apple's AirPlay 2 across its current product lineup. That's all on top of the whole-home audio platform that Sonos has built its brand on. But to stay so nimble and adaptable, Sonos constantly needs to be looking forward and have some sense of what's coming down the technology pike.