Aiming to represent user characteristics and personal interests, the task of profiling is playing an increasingly important role for many real-world applications, e.g., e-commerce social networks platforms. By exploiting data like texts behaviors, most existing solutions address as a classification task, where each formulated individual instance. Nevertheless, user's profile not only reflected from her/his affiliated data, but also can be inferred other users, users that have similar...

Hierarchical user profiling that aims to model users' real-time interests in different granularity is an essential issue for personalized recommendations E-commerce. On one hand, items (i.e. products) are usually organized hierarchically categories, and correspondingly naturally hierarchical on of categories. the other multiple oriented become very popular E-commerce sites, which require as well. In this paper, we propose HUP, a User Profiling framework solve problem recommender systems....

Recommender Systems have been playing essential roles in e-commerce portals. Existing recommendation algorithms usually learn the ranking scores of items by optimizing a single task (e.g. Click-through rate prediction) based on users' historical click sequences, but they generally pay few attention to simultaneously modeling multiple types behaviors or jointly optimize objectives both and Conversion rate), which are vital for sites. In this paper, we argue that it is crucial formulate...

Inferring substitutable and complementary items is an important fundamental concern for recommendation in e-commerce websites. However, the item relationships real-world are usually heterogeneous, posing great challenges to conventional methods that can only deal with homogeneous relationships. More specifically, this problem, there a lack of in-depth investigation on 1) decoupling semantics modeling heterogeneous relationships, at same time, 2) incorporating mutual influence between...

In this paper, we study collaborative filtering in an interactive setting, which the recommender agents iterate between making recommendations and updating user profile based on feedback. The most challenging problem scenario is how to suggest items when has not been well established, \ie recommend for cold-start users or warm-start with taste drifting. Existing approaches either rely overly pessimistic linear exploration strategy adopt meta-learning algorithms a full exploitation way. work,...

Modeling user's historical feedback is essential for Click-Through Rate Prediction in personalized search and recommendation. Existing methods usually only model users' positive information such as click sequences which neglects the context of feedback. In this paper, we propose a new perspective context-aware behavior modeling by including whole page-wisely exposed products corresponding contextualized page-wise sequence. The intra-page inter-page interest evolution can be captured to learn...

The rapid spread of smartphones has led to the increasing popularity Location-Based Social Networks(LBSNs) like Foursquare, Gowalla, Facebook Places and so on where users can publish information about their current location. In LBSNs, identifying home locations is very important for various applications effective location-based advertisement recommendation. However, this problem rather challenging because location in LBSNs sparse noisy: Only a small percentage share due privacy concerns; may...

Multi-scenario Learning to Rank is essential for Recommender Systems, Search Engines and Online Advertising in e-commerce portals where the ranking models are usually applied many scenarios. However, existing works mainly focus on learning model a single scenario, pay less attention multiple We identify two practical challenges industrial multi-scenario systems: (1) The Feedback Loop problem that always trained items chosen by ranker itself. (2) Insufficient training data small new To...

Event-based Social Networks(EBSNs) which combine online interactions and offline events among users have experienced increased popularity rapid growth recently. In EBSNs, event recommendation is significant for due to the extremely large amount of events. However, problem rather challenging because it faces a serious cold-start problem: Events short life time new are registered by only few users. What's more, there implicit feedback information. Existing approaches like collaborative...

The rapid spread of mobile internet and location-acquisition technologies have led to the increasing popularity Location-Based Social Networks(LBSNs). Users in LBSNs can share their life by checking at various venues any time. In LBSNs, identifying home locations users is significant for effective location-based services like personalized search, targeted advertisement, local recommendation so on. this paper, we propose a Home Location Global Positioning System called HLGPS tackle with...

Event sequence, where each event is associated with a marker and timestamp, increasingly ubiquitous in various applications. Accordingly, forecasting emerges to be crucial problem, which aims predict the next based on historical sequence. In this paper, we propose ANPP, an Attentive Neural Point Processes framework solve problem. comparison state-of-the-art methods like recurrent marked temporal point processes, ANPP leverages time-aware self-attention mechanism explicitly model influence...

Poly (lactic acid)/poly (glycolic acid) (LG) bone scaffold exhibits good biocompatibility for defect regeneration but lacks satisfactory mechanical and osteogenic induction properties. Here, graphene oxide (GO) was encapsulated by polydopamine (PDA) via self-polymerization of dopamine, strontium (Sr) loaded onto GO the chelation PDA. The modified added to LG prepared selective laser sintering as a reinforcing phase improve properties results indicated that tensile compressive strengths with...

Social Networks have experienced increased popularity and rapid growth in recent years. Recommendation is significant for users due to the extremely large amount of information Networks. Most existing recommender systems rely on collaborative filtering techniques which focus recommendi ng most relevant items based past rating or items. In Networks, cold-start data sparsity problems are very serious because new growing rapidly. Taking Event problem Event-Based as a scenario, many events newly...

Efficiently inducing high-level interpretable regularities from knowledge graphs (KGs) is an essential yet challenging task that benefits many downstream applications. In this work, we present GPFL, a probabilistic rule learner optimized to mine instantiated first-order logic rules KGs. Instantiated contain constants extracted Compared abstract no constants, are capable of explaining and expressing concepts in more details. GPFL utilizes novel two-stage generation mechanism first generalizes...

With the extensive use of sensor-embedded smart phones, Location-Based Social Networks (LBSN) become more and popular among online social networks in recent years. In networks, constructing shortest path with minimum cost between any two nodes efficiently is vital for both graph analysis implementation applications. This well known as routing problem networks. However, existing approaches all fail scenario LBSN which are large dynamic. this paper, we work out a fast system leveraging...

Consuming behaviors of users form sequences ordered by time intuitively. Long Short-Term Memory Based Recurrent Neural Networks(LSTM), which are special kind Networks, ideal for modeling sequences. In this work, we propose a LSTM based model called CF-LSTM can the consuming Collaborative Filtering(CF). To effectively train model, step-combine technique, processes k ratings at step and solves long problem ratings. improve performance CF-LSTM, extend our with ordinal cost considering ordinary...

Enormous efforts of human volunteers have made Wikipedia become a treasure textual knowledge. Relation extraction that aims at extracting structured knowledge in the unstructured texts is an appealing but quite challenging problem because it's hard for machines to understand plain texts. Existing methods are not effective enough they relation types level without exploiting behind In this paper, we propose novel framework called Athena leveraging Intrinsic Patterns which patterns can semantic...

Along with the rise of Event-Based Social Networks (EBSNs), event recommendation has become an increasing important problem. However, unlike recommending usual items, such as movies or music, suffers from severe cold-start problem, because most events in EBSNs are typically short-lived and registered by only a few users. Additionally, available feedbacks for implicit feedbacks. In this work, we propose Hybrid Deep Neural Collaborative Filtering Architecture (HDNN-CF) that collaboratively...

Abstract Spatial resolution of diffusion tensor images is usually compromised to accelerate the acquisitions, and state‐of‐the‐art (SOTA) image super‐resolution (SR) reconstruction methods are commonly based on supervised learning models. Considering that matched low‐resolution (LR) high‐resolution (HR) diffusion‐weighted (DW) pairs not readily available, we propose a semi‐supervised DW SR method multiple references (MRSR) extracted from other subjects. In MRSR, prior information HR...

Abstract HB(N 5 ) 3 M 1∼2 (N BH (M = Be, Mg, Ca, Zn, and Cd) have been investigated as potential high‐energy density materials in this article by means of functional theory. They are all content show kinetic stability to the breakup N ring, especially Mg(N BH. The MN bondings mononuclear compounds stronger than those corresponding dinuclear ones, except Be analogues. Furthermore, Zn 2 BH, shown be energetically stable with respect disproportionation. © 2009 Wiley Periodicals, Inc. Int J...