Knowledge Graph (KG) reasoning has been an interesting topic in recent decades. Most current researches focus on predicting the missing facts for incomplete KG. Nevertheless, Temporal KG (TKG) reasoning, which is to forecast future facts, still faces with a dilemma due complex interactions between entities over time. This article proposes novel intricate Spatiotemporal Dependency learning Network (STDN) based Convolutional (GCN) capture underlying correlations of entity at different...

Document-level Relation Extraction (RE) is particularly challenging due to complex semantic interactions among multiple entities in a document. Among exiting approaches, Graph Convolutional Networks (GCN) one of the most effective approaches for document-level RE. However, traditional GCN simply takes word nodes and adjacency matrix represent graphs, which difficult establish direct connections between distant entity pairs. In this paper, we propose Global Context-enhanced (GCGCN), novel...

The Short Message Service (SMS) has widely extended in the modern methods of communication technology. classification spam message is an interesting and prominent issue. Classifying availability SMS a challenging task, plenty research been carried out this direction employing Machine Learning techniques such as Naive Bayes (NB), Random Forest (RF), Support Vector (SVM) for Spam Classification. Although these have shown adequate performance, but are not efficient enough terms classification....

Analyzing omics data from a network-based perspective can facilitate biomarker discovery. To improve disease diagnosis and identify prospective information indicating the onset of complex disease, computational method for identifying potential biomarkers based on differential sub-networks (PB-DSN) is developed. In PB-DSN, Pearson correlation coefficient (PCC) used to measure relationship between feature ratios infer networks. A sub-network extracted crucial discriminating different groups...

Abstract Motivation Hypothesis generation (HG) refers to the discovery of meaningful implicit connections between disjoint scientific terms, which is great significance for drug discovery, prediction side effects and precision treatment. More recently, a few initial studies attempt model dynamic meaning terms or term pairs HG. However, most existing methods still fail accurately capture utilize evolution relations. Results This article proposes novel temporal difference embedding (TDE)...

This paper focus on video recommendation in response to the explosive growth of movies which has made people dilemma information overload for a long time. It is particularly important push accurately. Regarding this situation, studies how recommend most similar users based their interest tags and browsing behavior. Compared with collaborative filtering algorithm, whose processing effect not good confronting scoring data sparse problem. makes full use movie review data, experimental set...

In this paper, we propose a novel model called Adversarial Multi-Task Network (AMTN) for jointly modeling Recognizing Question Entailment (RQE) and medical Answering (QA) tasks. AMTN utilizes pre-trained BioBERT an Interactive Transformer to learn the shared semantic representations across different task through parameter sharing mechanism. Meanwhile, adversarial training strategy is introduced separate private features of each from representations. Experiments on BioNLP 2019 RQE QA Shared...

Analyzing the disease data from view of combinatorial features may better characterize phenotype. In this study, a novel method is proposed to construct feature combinations and classification model (CFC-CM) by mining key relationships. CFC-CM iteratively tests for differences in relationship between different groups. To do this, it uses modified $k$k-top-scoring pair (M-$k$k-TSP) algorithm then selects most discriminative pairs current set infer build model. Compared with support vector...

In systems biology, filtering the discriminative features from complex high-dimensional data is a crucial issue. This paper proposes feature selection algorithm based on overlapping and group (FS-FOGO) to calculate importance. FS-FOGO weighs two aspects: degree ratio of area effective range each class proportion heterogeneous samples in every sample's nearest neighbors. To show validation FS-FOGO, it compared with gene (ERGS), which calculates weights range, six public biological sets one...

Decision-making is a basic component of agents' (e.g., intelligent sensors) behaviors, in which one's cognition plays crucial role the process and outcome. Extensive games, class interactive decision-making scenarios, have been studied diverse fields. Recently, model extensive games was proposed agent structure underlying game quality situations are encoded by artificial neural networks. This refines classic corresponding equilibrium concept-

Generating biomedical hypotheses is a difficult task as it requires uncovering the implicit associations between massive scientific terms from large body of published literature. A recent line Hypothesis Generation (HG) approaches - temporal graph-based have shown great success in modeling evolution term-pair relationships. However, these model each term or with Recurrent Neural Network (RNN) independently, which neglects rich covariation among all term-pairs while ignoring direct...

Hypothesis Generation (HG) aims to expedite biomedical researches by generating novel hypotheses from existing scientific literature. Most studies focused on modeling static snapshots of the corpus, neglecting temporal evolution terms. Despite recent efforts learn term Knowledge Bases (KBs) for HG, information multi-source KBs is still overlooked, which contains important, up-to-date knowledge. In this paper, an innovative Temporal Contrastive Learning (TCL) framework introduced uncover...