Existing item-based collaborative filtering (ICF) methods leverage only the relation of collaborative similarity. Nevertheless, there exist multiple relations between items in real-world scenarios. Distinct from the collaborative similarity that implies co-interact patterns from the user perspective, these relations reveal fine-grained knowledge on items from different perspectives of meta-data, functionality, etc. However, how to incorporate multiple item relations is less explored in recommendation research. In this work, we propose Relational Collaborative Filtering (RCF), a general framework to exploit multiple relations between items in recommender system. We find that both the relation type and the relation value are crucial in inferring user preference. To this end, we develop a two-level hierarchical attention mechanism to model user preference. The first-level attention discriminates which types of relations are more important, and the second-level attention considers the specific relation values to estimate the contribution of a historical item in recommending the target item. To make the item embeddings be reflective of the relational structure between items, we further formulate a task to preserve the item relations, and jointly train it with the recommendation task of preference modeling. Empirical results on two real datasets demonstrate the strong performance of RCF. Furthermore, we also conduct qualitative analyses to show the benefits of explanations brought by the modeling of multiple item relations.

Email has remained a principal form of communication among people, both in enterprise and social settings. With a deluge of emails crowding our mailboxes daily, there is a dire need of smart email systems that can recover important emails and make personalized recommendations. In this work, we study the problem of predicting user triage actions to incoming emails where we take the reply prediction as a working example. Different from existing methods, we formulate the triage action prediction as a recommendation problem and focus on the content-based approach, where the users are represented using the content of current and past emails. We also introduce additional similarity features to further explore the affinities between users and emails. Experiments on the publicly available Avocado email collection demonstrate the advantages of our proposed recommendation framework and our method is able to achieve better performance compared to the state-of-the-art deep recommendation methods. More importantly, we provide valuable insight into the effectiveness of different textual and user representations and show that traditional bag-of-words approaches, with the help from the similarity features, compete favorably with the more advanced neural embedding methods.

Mobile context determination is an important step for many context aware services such as location-based services, enterprise policy enforcement, building or room occupancy detection for power or HVAC operation, etc. Especially in enterprise scenarios where policies (e.g., attending a confidential meeting only when the user is in "Location X") are defined based on mobile context, it is paramount to verify the accuracy of the mobile context. To this end, two stochastic models based on the theory of Hidden Markov Models (HMMs) to obtain mobile context are proposed-personalized model (HPContext) and collaborative filtering model (HCFContext). The former predicts the current context using sequential history of the user's past context observations, the latter enhances HPContext with collaborative filtering features, which enables it to predict the current context of the primary user based on the context observations of users related to the primary user, e.g., same team colleagues in company, gym friends, family members, etc. Each of the proposed models can also be used to enhance or complement the context obtained from sensors. Furthermore, since privacy is a concern in collaborative filtering, a privacy-preserving method is proposed to derive HCFContext model parameters based on the concepts of homomorphic encryption. Finally, these models are thoroughly validated on a real-life dataset.

From suggesting the meaning of life is'chocolate' to stating that having a boyfriend just ends in heartbreak and loneliness, these conversations show just how sassy Apple's Siri can be. A hilarious gallery, collated by Bored Panda, shows the amusing automated responses owners of Apple gadgets have received when trying to communicate with their virtual assistant. Many people from around the world have contributed conversations in which Siri struggled to understand what they were trying to say. One person was asked by their virtual assistant if they have'anything better to do' after they sent it a tongue-twister. Another was informed by Siri that a reminder was set to tell them they're dumb.

There will be no fourth season for Netflix's dark comedy series Santa Clarita Diet as the streaming giant has opted to cancel the series starring Drew Barrymore and Timothy Olyphant. The news comes almost a month after the March 29 release of the show's third season. Santa Clarita Diet is the latest Netflix series canceled after three seasons. Following Deadline's story examining the trend, fans of the show started a #SaveSantaClaritaDiet Twitter campaign, rallying for a Season 4 renewal. Like with the recently axed Netflix comedy series One Day At a Time, also after three seasons, the cancellation for Santa Clarita Diet comes after a very strong season, which scored 100% on Rotten Tomatoes.

Garage door openers are certainly convenient, but compared to the connected devices around our homes, they're nor very smart. But it doesn't have to be that way. Today, you can choose from two smart garage door remotes at all-time low Amazon prices to beef up your existing system: the Chamberlain MyQ for $50Remove non-product link, down from a list price of $80, or the Nexx Garage for $70Remove non-product link, down from a list price of $100. The MyQ smart garage door hub connects to any compatible garage door opener and Wi-Fi to bring smarts to your existing garage setup. Once connected, you'll be able to use the app to open and close your door, get alerts when the door opens, closes, or is left open too long, and schedule automatic close times.

Some weeks, there are so many tall-skinny cylinders sitting around my apartment you might think I spend my time as a wizard in The Cones of Dunshire. The job of a WIRED Gear reviewer does not include making civilizations to collect cones. It does involve testing a ton of can-shaped portable speakers. The JBL Link 20 is one my favorites, and it's half off through April 27. Usually speakers come in, and I test them, and then they leave my cluttered life.

Forty-one percent of voice assistant users are concerned about trust, privacy and passive listening, according to a new report from Microsoft focused on consumer adoption of voice and digital assistants. And perhaps people should be concerned -- all the major voice assistants, including those from Google, Amazon, Apple and Samsung, as well as Microsoft, employ humans who review the voice data collected from end users. But people didn't seem to know that was the case. So when Bloomberg recently reported on the global team at Amazon that reviews audio clips from commands spoken to Alexa, some backlash occurred. In addition to the discovery that our AI helpers also have a human connection, there were concerns over the type of data the Amazon employees and contractors were hearing -- criminal activity and even assaults in a few cases, as well as the otherwise odd, funny or embarrassing things the smart speakers picked up.

Most modern successful recommender systems are based on matrix factorization techniques, i.e., learning a latent embedding for each user and each item from the given rating matrix and use the embeddings to complete the matrix. However, these learned latent embeddings are inherently transductive and are not designed to generalize to unseen users/items or new tasks. In this paper, we aim to learn an inductive model for recommender systems based on the local graph patterns around user-item pairs. The inductive model can generalize to unseen nodes/items, and potentially also transfer to other tasks. To learn such a model, we extract a local enclosing subgraph for each training (user, item) pair, and feed the subgraphs to a graph neural network (GNN) to train a rating prediction model. We show that our model achieves highly competitive performance with state-of-the-art transductive methods, and is more stable when the rating matrix is sparse. Furthermore, our transfer learning experiment validates that the learned model is transferrable to new tasks.

Poor lifestyle represents a health risk factor and is the leading cause of morbidity and chronic conditions. The impact of poor lifestyle can be significantly altered by individual behavior change. Although the current shift in healthcare towards a long-lasting modifiable behavior, however, with increasing caregiver workload and individuals' continuous needs of care, there is a need to ease caregiver's work while ensuring continuous interaction with users. This paper describes the design and validation of CoachAI, a conversational agent-assisted health coaching system to support health intervention delivery to individuals and groups. This research provides three main contributions to the preventive healthcare & healthy lifestyle promotion: (1) it presents the conversational agent to aid the caregiver; (2) it aims to decrease caregiver's workload and enhance care given to users, by handling (automating) repetitive caregiver tasks; and (3) it presents a domain-independent mobile health conversational agent for health intervention delivery. We will discuss our approach and analyze the results of a one-month validation study on physical activity, healthy diet and stress management. Introduction Adhering to a healthy lifestyle is among the most contributor to health promotion and disease prevention [27,28]. A varied diet and regular physical activity have significant benefits for individuals' overall health [27,28,6]. Similarly, mental wellness is associated with social competence and coping skills that lead to positive outcomes in adulthood and later stages of individuals life [29,9]. Although the benefit of pursuing a healthy lifestyle, several barriers exist in the process of health promotion. For instance, individuals' motivation to change, their demographics and preparedness are all factors that contribute to their intention to follow a healthy lifestyle. Several studies tackled the issue of poor lifestyle through mobile technologies. Approaches [31,32] developed mobile applications to mitigate the risk of poor diet, sedentary lifestyle and anxiety. That said, the learning curve associated with mobile apps is still an issue, especially for individuals with low digital literacy.