In the new'Child's Play' film, the wicked Chucky doll is reconceived as an A.I. assistant.

With people examining further into the universe of innovation, artificial intelligence or AI is presently ubiquitous. Interestingly, albeit artificial intelligence is an innovation of interest at this moment, the idea has been around for over seven-plus decades now. With increased awareness about the transformative intensity of AI, the adoption of AI is presently being seen across numerous enterprises. From enormous banks like HDFC Bank to e-commerce business players like Pepperfry, AI is being utilized for a large variety of use cases. Eyeballing is of fundamental significance in the credit decision procedure, and the bot emulates a human by eyeballing the client demographics for the individuals who have applied for advances against the already existing base.

This paper studies recommender systems with knowledge graphs, which can effectively address the problems of data sparsity and cold start. Recently, a variety of methods have been developed for this problem, which generally try to learn effective representations of users and items and then match items to users according to their representations. Though these methods have been shown quite effective, they lack good explanations, which are critical to recommender systems. In this paper, we take a different path and propose generating recommendations by finding meaningful paths from users to items. Specifically, we formulate the problem as a sequential decision process, where the target user is defined as the initial state, and the walks on the graphs are defined as actions. We shape the rewards according to existing state-of-the-art methods and then train a policy function with policy gradient methods. Experimental results on three real-world datasets show that our proposed method not only provides effective recommendations but also offers good explanations .

Building multi-domain AI agents is a challenging task and an open problem in the area of AI. Within the domain of dialog, the ability to orchestrate multiple independently trained dialog agents, or skills, to create a unified system is of particular significance. In this work, we study the task of online posterior dialog orchestration, where we define posterior orchestration as the task of selecting a subset of skills which most appropriately answer a user input using features extracted from both the user input and the individual skills. To account for the various costs associated with extracting skill features, we consider online posterior orchestration under a skill execution budget. We formalize this setting as Context Attentive Bandit with Observations (CABO), a variant of context attentive bandits, and evaluate it on simulated non-conversational and proprietary conversational datasets.

Appliances, even large ones, are getting in on the action. But that doesn't mean you need to toss your old "dumb" stuff and buy brand new smart devices. Air conditioners are certainly one type of appliance that is heading down the smart path, but if you have a window or wall unit you already love (or can't afford to replace), you might consider turning your dumb unit into a smart A/C. With a few tweaks, you can convert the unit you have into a cooling solution you can turn on and off remotely with an app or through your favorite smart home ecosystems, such as Google Assistant or Alexa. A window or wall unit (or a portable A/C) works just fine when you manually turn it on with buttons, so why is it even necessary to make it smart?

We trust smart assistants like Alexa, Siri, and Google Assistant to control so many things in our homes--door locks, lighting, security cameras--but should we allow them to light real-flame candles with a simple command? That's the question I found myself pondering while I was testing the Ludela Perfect Pillar smart candle, and the answer I settled on was, "mostly, no." The Ludela Perfect Pillar is a remote-controlled candle designed with safety in mind. It ignites using a canister of liquid paraffin, but only when a specific sequence of buttons is pressed on the remote. The sequence is simple to remember but hard to guess, so I didn't feel uncomfortable leaving the remote within my children's reach.

It's tough to do much of anything involving technology these days without running into a virtual assistant. Pick up your Android phone or Chromebook, and there's Google Assistant waiting for a chat. Apple's got Siri, poor Samsung's got Bixby, and even random companies like Bank of America are getting in on the action with their own woefully unnecessary A.I. personalities (sorry, "Erica"). We've talked plenty about the reasons why everyone and their mother wants you to get friendly with their flavor of robot aid -- and why that, in turn, has led to what I call the post-OS era, in which a device's operating system is less important than the virtual assistant threaded throughout it. It's no coincidence that Google is slowly expanding Assistant into a platform of its own, and what we're seeing now is almost certainly just the tip of the iceberg.

In the race among tech companies to bring their voice recognition technology into the realm of personal medicine, Google is the furthest along, according to a study published on Thursday in the journal Nature Digital Medicine. Researchers Yan Fossat and Adam Palanica from lab company Klick Health in Toronto tested technology from Google, Amazon and Apple to gauge how well their services comprehended the 50 most commonly prescribed medicines and whether they could provide accurate information to users. Fossat and Palanica said they activated Google Assistant, Amazon Alexa and Apple's Siri and played individual audio clips from 46 English speaking people with the prompt, "Tell me about," followed by the medication name. "We reviewed all the literature, and identified this one area of medication comprehension that is under studied," Fossat told CNBC. "It's especially important to research these voice assistant tools, given the growing demand for them in health care."

Embedding models, which learn latent representations of users and items based on user-item interaction patterns, are a key component of recommendation systems. In many applications, contextual constraints need to be applied to refine recommendations, e.g. when a user specifies a price range or product category filter. The conventional approach, for both context-aware and standard models, is to retrieve items and apply the constraints as independent operations. The order in which these two steps are executed can induce significant problems. For example, applying constraints a posteriori can result in incomplete recommendations or low-quality results for the tail of the distribution (i.e., less popular items). As a result, the additional information that the constraint brings about user intent may not be accurately captured. In this paper we propose integrating the information provided by the contextual constraint into the similarity computation, by merging constraint application and retrieval into one operation in the embedding space. This technique allows us to generate high-quality recommendations for the specified constraint. Our approach learns constraints representations jointly with the user and item embeddings. We incorporate our methods into a matrix factorization model, and perform an experimental evaluation on one internal and two real-world datasets. Our results show significant improvements in predictive performance compared to context-aware and standard models.

Mobile sequential recommendation was originally designed to find a promising route for a single taxicab. Directly applying it for multiple taxicabs may cause an excessive overlap of recommended routes. The multi-taxicab recommendation problem is challenging and has been less studied. In this paper, we first formalize a collective mobile sequential recommendation problem based on a classic mathematical model, which characterizes time-varying influence among competing taxicabs. Next, we propose a new evaluation metric for a collection of taxicab routes aimed to minimize the sum of potential travel time. We then develop an efficient algorithm to calculate the metric and design a greedy recommendation method to approximate the solution. Finally, numerical experiments show the superiority of our methods. In trace-driven simulation, the set of routes recommended by our method significantly outperforms those obtained by conventional methods.