Text Summarization models facilitate biomedical clinicians and researchers in acquiring informative data from enormous domain-specific literature within less time and effort.Evaluating and selecting the most informati...Text Summarization models facilitate biomedical clinicians and researchers in acquiring informative data from enormous domain-specific literature within less time and effort.Evaluating and selecting the most informative sentences from biomedical articles is always challenging.This study aims to develop a dual-mode biomedical text summarization model to achieve enhanced coverage and information.The research also includes checking the fitment of appropriate graph ranking techniques for improved performance of the summarization model.The input biomedical text is mapped as a graph where meaningful sentences are evaluated as the central node and the critical associations between them.The proposed framework utilizes the top k similarity technique in a combination of UMLS and a sampled probability-based clustering method which aids in unearthing relevant meanings of the biomedical domain-specific word vectors and finding the best possible associations between crucial sentences.The quality of the framework is assessed via different parameters like information retention,coverage,readability,cohesion,and ROUGE scores in clustering and non-clustering modes.The significant benefits of the suggested technique are capturing crucial biomedical information with increased coverage and reasonable memory consumption.The configurable settings of combined parameters reduce execution time,enhance memory utilization,and extract relevant information outperforming other biomedical baseline models.An improvement of 17%is achieved when the proposed model is checked against similar biomedical text summarizers.展开更多
The structural,electronic and optical properties of KNbO 3 （KN）,NaNbO3（NN）and K05 Na0.5NbO3（KNN） in paraelectric cubic phase were calculated employing the plane-wave pseudopotential method based on density funct...The structural,electronic and optical properties of KNbO 3 （KN）,NaNbO3（NN）and K05 Na0.5NbO3（KNN） in paraelectric cubic phase were calculated employing the plane-wave pseudopotential method based on density functional theory （DFT）.The calculated electronic structures of the three crystals show similar features in the valence bands and the lower conduction bands.However,the structures in higher conduction bands differ markedly due to the effect of Na and K atoms.The calculated optical properties reveal that the features of optical spectrum at low energy are dominated by the transitions from O2p valence bands to Nb 4d conduction bands and those at high energy are related to the transitions to K 4s4p and/or Na 3s3p states.Moreover,the optical constants of KNN are approximately the average of KN and NN at high energy.Therefore,the optical properties of KNN in high energy region can probably be altered by changing the ratio of Na/K.展开更多
An important challenge facing K-ion batteries lies in exploring earth-abundant and safe cathode materials that can provide high capacity with high migration rate of K ions.Here,we propose a simple and efficient method...An important challenge facing K-ion batteries lies in exploring earth-abundant and safe cathode materials that can provide high capacity with high migration rate of K ions.Here,we propose a simple and efficient method for searching potential K cathode materials with first principles calculations.Our screening is based on combinations of weight capacity,K ion occupation ratio,volume change per K,and valence limit.With this screening method we predicted a series of potential K ions cathodes with favorable electrochemical performance,such as K_(2)VPO_(4)CO_(3)-like structures with 1 D diffusion channels,3 D channel structures K_(2)CoSiO_(4),layered materials KCoO_(2),KCrO_(2),KVF_(4) and K_(5)V_(3)F_(14),and others.These potential cathodes have small volume changes,suitable voltage,and high capacity,with small diffusion barriers.They may be useful in K-ion batteries with high energy density and rate performance.展开更多
文摘Text Summarization models facilitate biomedical clinicians and researchers in acquiring informative data from enormous domain-specific literature within less time and effort.Evaluating and selecting the most informative sentences from biomedical articles is always challenging.This study aims to develop a dual-mode biomedical text summarization model to achieve enhanced coverage and information.The research also includes checking the fitment of appropriate graph ranking techniques for improved performance of the summarization model.The input biomedical text is mapped as a graph where meaningful sentences are evaluated as the central node and the critical associations between them.The proposed framework utilizes the top k similarity technique in a combination of UMLS and a sampled probability-based clustering method which aids in unearthing relevant meanings of the biomedical domain-specific word vectors and finding the best possible associations between crucial sentences.The quality of the framework is assessed via different parameters like information retention,coverage,readability,cohesion,and ROUGE scores in clustering and non-clustering modes.The significant benefits of the suggested technique are capturing crucial biomedical information with increased coverage and reasonable memory consumption.The configurable settings of combined parameters reduce execution time,enhance memory utilization,and extract relevant information outperforming other biomedical baseline models.An improvement of 17%is achieved when the proposed model is checked against similar biomedical text summarizers.
基金Supported by the National Natural Science Foundation of China (Nos. 50862005,51062005,91022034 and 91022027)the Jiangxi Natural Science Foundation and Cooperative Project (Nos. 2008GZC000,2009JX02060,2010GQW0038 and [2008]212)+1 种基金Foundation of Jiangxi Educational Committee (GJJ11204)the Jiangxi Colleges and Universities "Advanced Ceramics" scientific and technological innovation team
文摘The structural,electronic and optical properties of KNbO 3 （KN）,NaNbO3（NN）and K05 Na0.5NbO3（KNN） in paraelectric cubic phase were calculated employing the plane-wave pseudopotential method based on density functional theory （DFT）.The calculated electronic structures of the three crystals show similar features in the valence bands and the lower conduction bands.However,the structures in higher conduction bands differ markedly due to the effect of Na and K atoms.The calculated optical properties reveal that the features of optical spectrum at low energy are dominated by the transitions from O2p valence bands to Nb 4d conduction bands and those at high energy are related to the transitions to K 4s4p and/or Na 3s3p states.Moreover,the optical constants of KNN are approximately the average of KN and NN at high energy.Therefore,the optical properties of KNN in high energy region can probably be altered by changing the ratio of Na/K.
基金supported by the National Key R&D Program of China(Grant No.2016YFA0200400)the National Natural Science Foundation of China(Grant No.11504123 and No.51627805)。
文摘An important challenge facing K-ion batteries lies in exploring earth-abundant and safe cathode materials that can provide high capacity with high migration rate of K ions.Here,we propose a simple and efficient method for searching potential K cathode materials with first principles calculations.Our screening is based on combinations of weight capacity,K ion occupation ratio,volume change per K,and valence limit.With this screening method we predicted a series of potential K ions cathodes with favorable electrochemical performance,such as K_(2)VPO_(4)CO_(3)-like structures with 1 D diffusion channels,3 D channel structures K_(2)CoSiO_(4),layered materials KCoO_(2),KCrO_(2),KVF_(4) and K_(5)V_(3)F_(14),and others.These potential cathodes have small volume changes,suitable voltage,and high capacity,with small diffusion barriers.They may be useful in K-ion batteries with high energy density and rate performance.
