We propose a recommendation system that suggests treatment strategies to a therapist during the psychotherapy session in real-time. Our system uses a turn-level rating mechanism that predicts the therapeutic outcome by computing a similarity score between the deep embedding of a scoring inventory, and the current sentence that the patient is speaking. The system automatically transcribes a continuous audio stream and separates it into turns of the patient and of the therapist and perform real-time inference of their therapeutic working alliance. The dialogue pairs along with their computed working alliance as ratings are then fed into a deep reinforcement learning recommendation system where the sessions are treated as users and the topics are treated as items. Other than evaluating the empirical advantages of the core components on an existing dataset of psychotherapy sessions, we demonstrate the effectiveness of this system in a web app.

Do you know about Linkedin sponsored post specs? Which 740 million active users, LinkedIn has grown from a website for job seekers to a complete social media platform that professionals use even when they are not on the hunt. That's why a comprehensive content strategy on LinkedIn is crucial, especially if your business is B2B. In this article, we will teach you how to create consistent and compelling content for LinkedIn and use LinkedIn tools such as LinkedIn Sponsored Content Ads, LinkedIn event ads, LinkedIn image ads, Linkedin insight tag checker, and more so that your company Can attract business-minded followers and, eventually, customers. At its core, inbound marketing is about solving a problem for your potential customers, and a big part of that is where they are when they need help.

Some of the biggest deals in tech right now come courtesy of Samsung Week, a sales event running through November 1st, in which the electronics giant has discounted a wide swath of their current-model devices. Google dropped some compelling offers of their own, with a third off the Pixel 6a and 25 percent off their Pixel Buds Pro. We also found Halloween sales from JBL, XBOX and PlayStation with discounts on older models and titles. Amazon's Echo devices are also on sale, with the Echo, Echo Show 8 and Echo Show 5 all at least 40 percent off. Here are all the best tech deals from this week that you can still get today.

We recommend Amazon's Echo Show 5 for those looking to adopt a smarter type of alarm clock, and as of this writing Amazon is running a sale that brings two of the diminutive smart displays down to $60 when you use the code SHOW52PK at checkout. We've seen individual deals bring the device as low as $35 in recent months, so this offer represents a slightly better value if you know you want multiple displays around the house (or if you want to grab one for multiple people). The Echo Show 5 technically has an MSRP of $85, but its average street price has typically sat in the $40-50 range in recent months. Add two to your cart and use the code SHOW52PK at checkout to see this discount. We gave the Echo Show 5 a review score of 85 when it launched last year, praising its solid-for-its-size sound quality, compact design, and usual suite of Alexa-aided smart features. Its most natural home is on a bedside table, as the device is fitted with features like a "sunrise" alarm that gradually brightens its screen to more gently wake you up in the morning and the ability to tap the top of the display to snooze an alarm.

After plans to break physical barriers with his metaverse initiative, Meta CEO Mark Zuckerberg revealed plans for another globe-spanning artificial intelligence (AI) project earlier this year, this time a universal translation tool unlike any other. At the same time, the company that made itself famous (and notorious) for its social media networks also introduced another AI-powered tool, a virtual assistant. Both of these intelligent applications were intended to have practical use cases in Zuckerberg's metaverse, those were their intended uses but they will also have wider business applications that Meta is all too aware of. AI virtual assistants, of course, are already in wider use by organizations as chatbots to handle basic customer requests and interactions across a variety of digital services– including Meta's own popular platforms like Facebook Messenger, Instagram, and WhatsApp Business. The other, less well-known AI use case(s) is the language and translation exercises that provide alternatives to relying on human translators to provide accurate, expert-quality translations in real-time.

Conversica Unveils Powerfully Human™ AI Capabilities That Transform Marketing, Sales, and Customer Success Teams and Make Bots Obsolete New Generation Revenue Digital Assistants™ Feature Major Advancements Built to Uniquely Fuel Revenue Growth; Scale Conversational Engagement with Real-Life, Two-Way Interactions FOSTER CITY, Calif., October 27, 2022 – Conversica, Inc., the leading provider of conversation automation solutions for enterprise revenue teams, today unveils the most humanlike AI advancements available to its Conversational AI platform. The new Conversica Chat and Conversica Answers enable the company's Revenue Digital Assistants (RDAs) to engage every lead, prospect, and customer in unscripted conversations with authentic, AI-generated web chat, SMS, and email. For the first time, revenue teams are able to trust RDAs to execute business objectives autonomously and work alongside their human counterparts, increasing capacity and freeing teams to focus on strategic  activities while simultaneously delivering stellar customer experiences and dramatically growing revenue opportunities. A demonstration of the announcement will be simulcast today at 11 am PST at the W Hotel in San Francisco. Conversica's new generation of Revenue Digital Assistants arrives at a pivotal time for businesses. With company budgets tightening and customer expectations for brand interactions continuously increasing, revenue teams across industries are required to do more with less.  "As the economy continues to change, the old growth model—working one customer at a time—no longer works," said Jim Kaskade, CEO of Conversica. "Businesses simply can't staff to the level needed for one-to-one conversations with every prospect, lead, and existing customer. And today's buyers are too savvy to be moved by the many one-way message blasts or, worse, scripted bots that are painfully programmed to route frustrated customers to an already overwhelmed human.” Unlike scripted bots that force people down a predefined path, Conversica’s AI-powered RDAs leverage the most advanced, largest language models available to quickly interpret open-ended questions and dynamically generate responses just like a human would. With over a decade of expertise in AI that benefits from being trained by billions of revenue-centric interactions, Conversica’s RDAs have evolved and learned how to influence and persuade customers and prospects throughout buyers’ lifecycles. Customers benefit from solutions that offer an average of 24x return on investment, 40-50% conversion rates, and 5x more pipeline, at reduced costs to the business. “The impact of Conversica’s RDAs goes way beyond our expectations, and the new advancements up the ante,” said David Shell, Director of Global Omnichannel Demand Marketing at Rockwell Automation, a Conversica customer. “The potential to increase productivity, lower costs, accelerate the velocity of leads in the pipeline and multiply staff efforts make Conversica’s solutions vital for fueling revenue growth.” New advancements include: Conversica Chat that is powered by an advanced natural language processing (NLP) solution that engages web visitors in the moment through dynamically generated dialog which goes far beyond the scripted workflows of simple chatbots. RDAs autonomously answer open-ended questions, connect visitors with the right resources, and set up demos or follow-up meetings to unlock revenue opportunities. No programming is required, which makes this solution uniquely game-changing. In the future, this same NLP technology will be applied across channels, including voice. Conversica Answers enable Revenue Digital Assistants to learn customers’ businesses FAQ automatically, so they can autonomously and authentically answer questions across any channel, in any language, at any point in the customer journey where businesses want strict governance around the answers, such as in highly regulated industries or for companies with very specific brand guidelines.  Conversica Premium Skills power Revenue Digital Assistants with the knowledge to act like a human member of the team. Premium Skills enable RDAs to leverage sophisticated segmentation and insights from across the enterprise to deliver more highly personalized customer experiences. These skills have enriched content from insight platforms like Account Based Marketing, Customer Data Platforms, and even CRM platforms with recommendation capabilities like Salesforce’s Einstein.  “Companies must deliver personalized, back-and-forth, human-like conversations to their contacts at every point in the customer journey. Otherwise, they’re leaving revenue on the table,” Kaskade said. “Conversica's RDAs feature a new level of Powerfully Human exchanges powered by the most advanced conversational AI technology on the market today. One can no longer tell the difference between a human and a Conversica RDA. Welcome to the new era of business where no revenue opportunity is missed!" To view the live simulcast, go to: https://conversica.brandlive.com/Conversica-Revenue-Revolution/en  About Conversica Conversica’s Revenue Digital Assistants™ (RDAs) supercharge workforces such as marketing, sales, and customer success teams to acquire untapped revenue through perfectly structured conversations. With billions of human interactions spanning more than a decade, Conversica’s RDAs have learned to influence and persuade customers and prospects throughout the customer journey lifecycle. Unlike chatbots, Conversica RDAs are Powerfully Human and can hold meaningful conversations at every touchpoint to create brand loyalty and maximize every revenue opportunity. The result is increased operational efficiencies, reduced costs, and improved customer experiences. To learn more, visit conversica.com and follow the company on Twitter, LinkedIn and Facebook.   Conversica Media Contact:  Meriane Morselli, on behalf of Conversica  conversica@msrcommunications.com (415) 989-9000 ###

Large language models produce human-like text that drives a growing number of applications. However, recent literature and, increasingly, real world observations, have demonstrated that these models can generate language that is toxic, biased, untruthful or otherwise harmful. Though work to evaluate language model harms is under way, translating foresight about which harms may arise into rigorous benchmarks is not straightforward. To facilitate this translation, we outline six ways of characterizing harmful text which merit explicit consideration when designing new benchmarks. We then use these characteristics as a lens to identify trends and gaps in existing benchmarks. Finally, we apply them in a case study of the Perspective API, a toxicity classifier that is widely used in harm benchmarks. Our characteristics provide one piece of the bridge that translates between foresight and effective evaluation.

How can we recommend existing bundles to users accurately? How can we generate new tailored bundles for users? Recommending a bundle, or a group of various items, has attracted widespread attention in e-commerce owing to the increased satisfaction of both users and providers. Bundle matching and bundle generation are two representative tasks in bundle recommendation. The bundle matching task is to correctly match existing bundles to users while the bundle generation is to generate new bundles that users would prefer. Although many recent works have developed bundle recommendation models, they fail to achieve high accuracy since they do not handle heterogeneous data effectively and do not learn a method for customized bundle generation. In this paper, we propose BundleMage, an accurate approach for bundle matching and generation. BundleMage effectively mixes user preferences of items and bundles using an adaptive gate technique to achieve high accuracy for the bundle matching. BundleMage also generates a personalized bundle by learning a generation module that exploits a user preference and the characteristic of a given incomplete bundle to be completed. BundleMage further improves its performance using multi-task learning with partially shared parameters. Through extensive experiments, we show that BundleMage achieves up to 6.6% higher nDCG in bundle matching and 6.3x higher nDCG in bundle generation than the best competitors. We also provide qualitative analysis that BundleMage effectively generates bundles considering both the tastes of users and the characteristics of target bundles.

Click-Through Rate (CTR) prediction on cold users is a challenging task in recommender systems. Recent researches have resorted to meta-learning to tackle the cold-user challenge, which either perform few-shot user representation learning or adopt optimization-based meta-learning. However, existing methods suffer from information loss or inefficient optimization process, and they fail to explicitly model global user preference knowledge which is crucial to complement the sparse and insufficient preference information of cold users. In this paper, we propose a novel and efficient approach named RESUS, which decouples the learning of global preference knowledge contributed by collective users from the learning of residual preferences for individual users. Specifically, we employ a shared predictor to infer basis user preferences, which acquires global preference knowledge from the interactions of different users. Meanwhile, we develop two efficient algorithms based on the nearest neighbor and ridge regression predictors, which infer residual user preferences via learning quickly from a few user-specific interactions. Extensive experiments on three public datasets demonstrate that our RESUS approach is efficient and effective in improving CTR prediction accuracy on cold users, compared with various state-of-the-art methods.

Recommender systems are the primary interface connecting users to a wide variety of online content, and therefore must overcome a number of challenges across the user population in order to serve them equitably. To this end, in 2019 we released RecSim, a configurable platform for authoring simulation environments to facilitate the study of RL algorithms (the de facto standard ML approach for addressing sequential decision problems) in recommender systems. However, as the technology has progressed, it has become increasingly important to address the gap between simulation and real-world applications, ensuring that models are flexible and easily extendible, enabling probabilistic inference of user dynamics, and addressing computational efficiency. To address these issues, we recently released RecSim NG, the "Next Generation" of simulators for recommender systems research and development. RecSim NG is a response to a set of use cases that have emerged as important challenges in the application of simulation to real-world problems.