To solve the problem of missing many valid triples in knowledge graphs(KGs),a novel model based on a convolutional neural network(CNN)called ConvKG is proposed,which employs a joint learning strategy for knowledge gra...To solve the problem of missing many valid triples in knowledge graphs(KGs),a novel model based on a convolutional neural network(CNN)called ConvKG is proposed,which employs a joint learning strategy for knowledge graph completion(KGC).Related research work has shown the superiority of convolutional neural networks(CNNs)in extracting semantic features of triple embeddings.However,these researches use only one single-shaped filter and fail to extract semantic features of different granularity.To solve this problem,ConvKG exploits multi-shaped filters to co-convolute on the triple embeddings,joint learning semantic features of different granularity.Different shaped filters cover different sizes on the triple embeddings and capture pairwise interactions of different granularity among triple elements.Experimental results confirm the strength of joint learning,and compared with state-of-the-art CNN-based KGC models,ConvKG achieves the better mean rank(MR)and Hits@10 metrics on dataset WN18 RR,and the better MR on dataset FB15k-237.展开更多
Knowledge tracing is the key component in online individualized learning,which is capable of assessing the users'mastery of skills and predicting the probability that the users can solve specific problems.Availabl...Knowledge tracing is the key component in online individualized learning,which is capable of assessing the users'mastery of skills and predicting the probability that the users can solve specific problems.Available knowledge tracing models have the problem that the assessments are not directly used in the predictions.To make full use of the assessments during predictions,a novel model,named deep knowledge tracing embedding neural network(DKTENN),is proposed in this work.DKTENN is a synthesis of deep knowledge tracing(DKT)and knowledge graph embedding(KGE).DKT utilizes sophisticated long short-term memory(LSTM)to assess the users and track the mastery of skills according to the users'interaction sequences with skill-level tags,and KGE is applied to predict the probability on the basis of both the embedded problems and DKT's assessments.DKTENN outperforms performance factors analysis and the other knowledge tracing models based on deep learning in the experiments.展开更多
In recent years, West Africa has been confronted with hydro-climatic disasters causing crises in both urban and rural areas. The tragedy in the occurrence of such events lies in the recurrent aspect of high water and ...In recent years, West Africa has been confronted with hydro-climatic disasters causing crises in both urban and rural areas. The tragedy in the occurrence of such events lies in the recurrent aspect of high water and associated floods. The devastating floods observed in Africa’s major rivers have revealed the need to understand the causes of these phenomena and to predict their behavior in order to improve the safety of exposed people and property. The aim of this study is to reproduce flood flows using the GR4J (Rural Engineering Four Daily Parameters) model to analyze flood risk in the Oti watershed in Togo. Daily data on flows (m3/s), potential evapotranspiration (mm/day) and average precipitation (mm) over the basin from 1961-2022 collected at the National Meteorological Agency of Togo (ANAMET) and the Department of Water Resources in Lome, were used with the R software package airGR. The Data from the West African Cordex program from 1961-2100 were used to analyze projected flows. The results obtained show the GR4J model’s effectiveness in reproducing flood flows, indicating that observed flows are well simulated during the calibration and validation periods, with KGE values ranging from 0.73 to 0.85 at calibration and 0.62 to 0.81 at validation. These KGE values reflect the good performance of the GR4J model in simulating flood flows in the watershed. However, a deterioration in the KGE value was observed over the second validation period. Under these conditions, there may be false or missed alerts for flood prediction, and the use of this model should be treated with the utmost caution for decision-support purposes.展开更多
In this paper, the Klein-Gordon equation （KGE） with power law nonlinearity will be considered. The bifurcation analysis as well as the ansatz method of integration will be applied to extract soliton and other wave s...In this paper, the Klein-Gordon equation （KGE） with power law nonlinearity will be considered. The bifurcation analysis as well as the ansatz method of integration will be applied to extract soliton and other wave solutions. In particular, the second approach to integration will lead to a singular soliton solution. However, the bifurcation analysis will reveal several other solutions that are of prime importance in relativistic quantum mechanics where this equation appears.展开更多
基金Supported by the National Natural Science Foundation of China(No.61876144)。
文摘To solve the problem of missing many valid triples in knowledge graphs(KGs),a novel model based on a convolutional neural network(CNN)called ConvKG is proposed,which employs a joint learning strategy for knowledge graph completion(KGC).Related research work has shown the superiority of convolutional neural networks(CNNs)in extracting semantic features of triple embeddings.However,these researches use only one single-shaped filter and fail to extract semantic features of different granularity.To solve this problem,ConvKG exploits multi-shaped filters to co-convolute on the triple embeddings,joint learning semantic features of different granularity.Different shaped filters cover different sizes on the triple embeddings and capture pairwise interactions of different granularity among triple elements.Experimental results confirm the strength of joint learning,and compared with state-of-the-art CNN-based KGC models,ConvKG achieves the better mean rank(MR)and Hits@10 metrics on dataset WN18 RR,and the better MR on dataset FB15k-237.
文摘Knowledge tracing is the key component in online individualized learning,which is capable of assessing the users'mastery of skills and predicting the probability that the users can solve specific problems.Available knowledge tracing models have the problem that the assessments are not directly used in the predictions.To make full use of the assessments during predictions,a novel model,named deep knowledge tracing embedding neural network(DKTENN),is proposed in this work.DKTENN is a synthesis of deep knowledge tracing(DKT)and knowledge graph embedding(KGE).DKT utilizes sophisticated long short-term memory(LSTM)to assess the users and track the mastery of skills according to the users'interaction sequences with skill-level tags,and KGE is applied to predict the probability on the basis of both the embedded problems and DKT's assessments.DKTENN outperforms performance factors analysis and the other knowledge tracing models based on deep learning in the experiments.
文摘In recent years, West Africa has been confronted with hydro-climatic disasters causing crises in both urban and rural areas. The tragedy in the occurrence of such events lies in the recurrent aspect of high water and associated floods. The devastating floods observed in Africa’s major rivers have revealed the need to understand the causes of these phenomena and to predict their behavior in order to improve the safety of exposed people and property. The aim of this study is to reproduce flood flows using the GR4J (Rural Engineering Four Daily Parameters) model to analyze flood risk in the Oti watershed in Togo. Daily data on flows (m3/s), potential evapotranspiration (mm/day) and average precipitation (mm) over the basin from 1961-2022 collected at the National Meteorological Agency of Togo (ANAMET) and the Department of Water Resources in Lome, were used with the R software package airGR. The Data from the West African Cordex program from 1961-2100 were used to analyze projected flows. The results obtained show the GR4J model’s effectiveness in reproducing flood flows, indicating that observed flows are well simulated during the calibration and validation periods, with KGE values ranging from 0.73 to 0.85 at calibration and 0.62 to 0.81 at validation. These KGE values reflect the good performance of the GR4J model in simulating flood flows in the watershed. However, a deterioration in the KGE value was observed over the second validation period. Under these conditions, there may be false or missed alerts for flood prediction, and the use of this model should be treated with the utmost caution for decision-support purposes.
文摘In this paper, the Klein-Gordon equation （KGE） with power law nonlinearity will be considered. The bifurcation analysis as well as the ansatz method of integration will be applied to extract soliton and other wave solutions. In particular, the second approach to integration will lead to a singular soliton solution. However, the bifurcation analysis will reveal several other solutions that are of prime importance in relativistic quantum mechanics where this equation appears.
