--Collaborative filtering is one of the most popular techniques in designing recommendation systems, and its most representative model, matrix factorization, has been wildly used by researchers and the industry. However, this model suffers from the lack of interpretability and the item cold-start problem, which limit its reliability and practicability. In this paper, we propose an interpretable recommendation model called Multi-Matrix F actorization (MMF), which addresses these two limitations and achieves the state-of-the-art prediction accuracy by exploiting common attributes that are present in different items. In the model, predicted item ratings are regarded as weighted aggregations of attribute ratings generated by the inner product of the user latent vectors and the attribute latent vectors. MMF provides more fine grained analyses than matrix factorization in the following ways: attribute ratings with weights allow the understanding of how much each attribute contributes to the recommendation and hence provide interpretability; the common attributes can act as a link between existing and new items, which solves the item cold-start problem when no rating exists on an item. We evaluate the interpretability of MMF comprehensively, and conduct extensive experiments on real datasets to show that MMF outperforms state-of-the-art baselines in terms of accuracy. I NTRODUCTION In recent years, recommendation systems gain increasing interest by both researchers and the industry [1], [2]. The most popular recommendation systems are based on collaborative filtering (CF) technique, which provides recommendations based on other similar users' choice [3]. Matrix factorization (MF) is one of the most common collaborative filtering models, whose main idea is to learn user latent vectors and item latent vectors, so that the inner product of the two vectors can approximate the original matrix with the minimal approximation error. MF has advantages of simplicity and performing well in many domains, such as recommendation systems, computer vision and document clustering [4]-[7]. However, it suffers from two limitations.

For years, the Sonos Wi-Fi speakers have been beloved by music fans for the ease of listening to streaming music at home with great sound. Also, add the ability to add multiple speakers for even better sound without having to resort to stringing speaker wire all over. The one drawback: Sonos speakers can be pricey, topping off at $499 per speaker and even higher for the $699 TV Playbar. So good news consumers: The most affordable way to get into the Sonos system goes on sale Thursday. But you won't find the speakers at Best Buy, Amazon or any of the other retailers which usually stock Sonos products.

Artificial intelligence-based property management platform Askporter has won a further £1.5 million in funding from half a dozen venture capitalists. One of the backers, Woking-based Henley Investments, has used the announcement to claim that the property management sector is'inefficient and ripe for disruption'. Askporter, which still calls itself a start-up despite being three years old, is a proptech company that uses a digital platform including a'virtual assistant' app to help property management teams run buildings more efficiently. It claims to help "cut costs and reduce the time spent on tedious tasks, all the while providing first-class support to occupants". "Our success raising £1.5 million investment reflects wider changes sweeping the property industry," says AskPorter CEO Tom Shrive (pictured, left).

If you ever get the creepy feeling you're being monitored when you use your computer, smartphone or smart speaker, our guest Geoffrey Fowler is here to tell you you are. Fowler writes a consumer-oriented technology column for The Washington Post. He's been investigating the ways our browsers and phone apps harvest personal information about us even while we're sleeping. And he discovered that Amazon had kept four years' worth of recorded audio from his home, captured by his Alexa smart speaker, including family conversations about medications and a friend doing a business transaction. Geoffrey Fowler joined the Post in 2017 after 16 years with the Wall Street Journal, writing about consumer technology, Silicon Valley, national affairs and China. He writes his technology column from San Francisco. He spoke with FRESH AIR's Dave Davies. You have a recent column. The headline is "I Found Your Data. It's For Sale." What kind of personal data did you find available for sale on the Internet? GEOFFREY FOWLER: I found all kinds of things that normal people would consider secrets and that corporations spend a lot of money - millions and millions of dollars - to try to keep out of the hands of their competitors and criminals. I found people's flight records. I found people's records from their doctors prescribing them medications. I found people's tax documents that they were - thought they were only sharing with their tax preparer. And they were available with one click. I could have opened them up and downloaded them. And where did this data come from?

Difficult co-workers who defy authority are more likely to cheat on their partners, a new study suggests. Researchers at the University of Texas discovered the correlation after studying the records of police officers, financial advisers, white-collar criminals and senior executives who used the Ashley Madison marital infidelity website. The data suggests a strong connection between people's actions in their personal and professional lives. They found that Ashley Madison were more than twice as likely to engage in corporate misconduct. Researchers investigated four study groups totalling 11,235 individuals.

Language systems have been of great interest to the research community and have recently reached the mass market through various assistant platforms on the web. Reinforcement Learning methods that optimize dialogue policies have seen successes in past years and have recently been extended into methods that personalize the dialogue, e.g. take the personal context of users into account. These works, however, are limited to personalization to a single user with whom they require multiple interactions and do not generalize the usage of context across users. This work introduces a problem where a generalized usage of context is relevant and proposes two Reinforcement Learning (RL)-based approaches to this problem. The first approach uses a single learner and extends the traditional POMDP formulation of dialogue state with features that describe the user context. The second approach segments users by context and then employs a learner per context. We compare these approaches in a benchmark of existing non-RL and RL-based methods in three established and one novel application domain of financial product recommendation. We compare the influence of context and training experiences on performance and find that learning approaches generally outperform a handcrafted gold standard.

Amazon staff review thousands of audio recordings made by Alexa each day -- including snippets of couples arguing and having sex -- an investigation claims. The clips were accidentally captured by the popular digital assistant -- confusing the noises for the commands it should be listening to -- and sent off for analysis. Staff at the tech firm review one in every five-hundred recordings made by Alexa, whether of deliberate commands to the assistant or accidental recordings. According to a privacy expert, the revelation is a reminder of the extent of the personal information that the tech firm has on its users. Amazon has an English-speaking team monitoring thousands of Alexa recordings daily based in Bucharest, Romania, the Sun claims, along with similar setups in Boston, Costa Rica and India.

This paper proposes a recommender system to alleviate the cold-start problem that can estimate user preferences based on only a small number of items. To identify a user's preference in the cold state, existing recommender systems, such as Netflix, initially provide items to a user; we call those items evidence candidates. Recommendations are then made based on the items selected by the user. Previous recommendation studies have two limitations: (1) the users who consumed a few items have poor recommendations and (2) inadequate evidence candidates are used to identify user preferences. We propose a meta-learning-based recommender system called MeLU to overcome these two limitations. From meta-learning, which can rapidly adopt new task with a few examples, MeLU can estimate new user's preferences with a few consumed items. In addition, we provide an evidence candidate selection strategy that determines distinguishing items for customized preference estimation. We validate MeLU with two benchmark datasets, and the proposed model reduces at least 5.92% mean absolute error than two comparative models on the datasets. We also conduct a user study experiment to verify the evidence selection strategy.

Now that it's upending the way you play music, cook, shop, hear the news and check the weather, the friendly voice emanating from your Amazon Alexa-enabled smart speaker is poised to wriggle its way into all things health care. Amazon has big ambitions for its devices. It thinks Alexa, the virtual assistant inside them, could help doctors diagnose mental illness, autism, concussions and Parkinson's disease. It even hopes Alexa will detect when you're having a heart attack. At present, Alexa can perform a handful of health care-related tasks: "She" can track blood glucose levels, describe symptoms, access post-surgical care instructions, monitor home prescription deliveries and make same-day appointments at the nearest urgent care center. Amazon has partnered with numerous health care companies, including several in California, to let consumers and employees use Alexa for health care purposes.

Dialogue systems are increasingly using knowledge bases (KBs) storing real-world facts to help generate quality responses. However, as the KBs are inherently incomplete and remain fixed during conversation, it limits dialogue systems' ability to answer questions and to handle questions involving entities or relations that are not in the KB. In this paper, we make an attempt to propose an engine for Continuous and Interactive Learning of Knowledge (CILK) for dialogue systems to give them the ability to continuously and interactively learn and infer new knowledge during conversations. With more knowledge accumulated over time, they will be able to learn better and answer more questions. Our empirical evaluation shows that CILK is promising.