Word sense disambiguation(WSD),identifying the specific sense of the target word given its context,is a fundamental task in natural language processing.Recently,researchers have shown promising results using long shor...Word sense disambiguation(WSD),identifying the specific sense of the target word given its context,is a fundamental task in natural language processing.Recently,researchers have shown promising results using long short term memory(LSTM),which is able to better capture sequential and syntactic features of text.However,this method neglects the dependencies among instances,such as their context semantic similarities.To solve this problem,we proposed a novel WSD model by introducing a cache-like memory module to capture the semantic dependencies among instances for WSD.Extensive evaluations on standard datasets demonstrate the superiority of the proposed model over various baselines.展开更多
Recycling of waste electrical and electronic equipment has become an urgent global issue in recent years from the prospectives of resources recycling and environmental protection.In the present work,the recycling of w...Recycling of waste electrical and electronic equipment has become an urgent global issue in recent years from the prospectives of resources recycling and environmental protection.In the present work,the recycling of waste memory modules(WMMs)through low-temperature alkali melts was investigated,based on the thermodynamic analysis of the nonmetallic reactions of brominated epoxy resin,glass fiber and memory chip with the molten mixed alkali.The effects of the reaction temperature and the ratio of alkali mixture on the removal rate of nonmetallic parts in WMMs were discussed under the condition of air atmosphere.The optimum process parameters were further confirmed by in-situ monitoring of the temperature during the whole reaction process.The mixtures with Cu,Fe and Ni as the main components were obtained after the treatment of WMMs in the molten alkali.These mixed metals were further separated into copper-rich and ferronickel-rich metals by physical magnetic separation.Moreover,the precious metals Au and Ag were enriched in Cu-rich alloys.This work provided an efficient and environment-friendly method for metal recycling from WMMs.展开更多
The widespread usage of Cyber Physical Systems(CPSs)generates a vast volume of time series data,and precisely determining anomalies in the data is critical for practical production.Autoencoder is the mainstream method...The widespread usage of Cyber Physical Systems(CPSs)generates a vast volume of time series data,and precisely determining anomalies in the data is critical for practical production.Autoencoder is the mainstream method for time series anomaly detection,and the anomaly is judged by reconstruction error.However,due to the strong generalization ability of neural networks,some abnormal samples close to normal samples may be judged as normal,which fails to detect the abnormality.In addition,the dataset rarely provides sufficient anomaly labels.This research proposes an unsupervised anomaly detection approach based on adversarial memory autoencoders for multivariate time series to solve the above problem.Firstly,an encoder encodes the input data into low-dimensional space to acquire a feature vector.Then,a memory module is used to learn the feature vector’s prototype patterns and update the feature vectors.The updating process allows partial forgetting of information to prevent model overgeneralization.After that,two decoders reconstruct the input data.Finally,this research uses the Peak Over Threshold(POT)method to calculate the threshold to determine anomalous samples from normal samples.This research uses a two-stage adversarial training strategy during model training to enlarge the gap between the reconstruction error of normal and abnormal samples.The proposed method achieves significant anomaly detection results on synthetic and real datasets from power systems,water treatment plants,and computer clusters.The F1 score reached an average of 0.9196 on the five datasets,which is 0.0769 higher than the best baseline method.展开更多
In order to implement semantic mapping of database metasearch engines, a system is proposed, which uses ontology as the organization form of information and records the new words not appearing in the ontology. When th...In order to implement semantic mapping of database metasearch engines, a system is proposed, which uses ontology as the organization form of information and records the new words not appearing in the ontology. When the new word＇ s frequency of use exceeds the threshold, it is added into the ontology. Ontology expansion is implemented in this way. The search process supports ＂and＂ and ＂or＂ Boolean operations accordingly. In order to improve the mapping speed of the system, a memory module is added which can memorize the recent query information of users and automatically learn the user＇ s query interest during the mapping which can dynamically decide the search order of instances tables. Experiments prove that these measures can obviously reduce the average mapping time.展开更多
Persistent indexing structures are proposed in response to emerging non-volatile memory(NVM)to provide high performance yet durable indexes.However,due to the lack of real NVM hardware,many prior persistent indexing s...Persistent indexing structures are proposed in response to emerging non-volatile memory(NVM)to provide high performance yet durable indexes.However,due to the lack of real NVM hardware,many prior persistent indexing structures were evaluated via emulation,which varies a lot across different setups and differs from the real deployment.Recently,Intel has released its Optane DC Persistent Memory Module(PMM),which is the first production-ready NVM.In this paper,we revisit popular persistent indexing structures on PMM and conduct comprehensive evaluations to study the performance differences among persistent indexing structures,including persistent hash tables and persistent trees.According to the evaluation results,we find that Cacheline-Conscious Extendible Hashing(CCEH)achieves the best performance among all evaluated persistent hash tables,and Failure-Atomic ShifT B+-Tree(FAST)and Write Optimal Radix Tree(WORT)perform better than other trees.Besides,we find that the insertion performance of hash tables is heavily influenced by data locality,while the insertion latency of trees is dominated by the flush instructions.We also uncover that no existing emulation methods accurately simulate PMM for all the studied data structures.Finally,we provide three suggestions on how to fully utilize PMM for better performance,including using clflushopt/clwb with sfence instead of clflush,flushing continuous data in a batch,and avoiding data access immediately after it is flushed to PMM.展开更多
文摘Word sense disambiguation(WSD),identifying the specific sense of the target word given its context,is a fundamental task in natural language processing.Recently,researchers have shown promising results using long short term memory(LSTM),which is able to better capture sequential and syntactic features of text.However,this method neglects the dependencies among instances,such as their context semantic similarities.To solve this problem,we proposed a novel WSD model by introducing a cache-like memory module to capture the semantic dependencies among instances for WSD.Extensive evaluations on standard datasets demonstrate the superiority of the proposed model over various baselines.
基金This work was supported by the National Natural Science Foundation of China(Grant Nos.51974288,51574216,51774264 and 51901231)the Liaoning Provincial Department of Education(Grant No.JYT19063)+1 种基金the Natural Science Foundation of Liaoning Province of China(Grant No.2019-MS-332)the Key Project of Innovation Foundation of IMR-CAS(Grant No.SCJJ-2013-ZD-03).
文摘Recycling of waste electrical and electronic equipment has become an urgent global issue in recent years from the prospectives of resources recycling and environmental protection.In the present work,the recycling of waste memory modules(WMMs)through low-temperature alkali melts was investigated,based on the thermodynamic analysis of the nonmetallic reactions of brominated epoxy resin,glass fiber and memory chip with the molten mixed alkali.The effects of the reaction temperature and the ratio of alkali mixture on the removal rate of nonmetallic parts in WMMs were discussed under the condition of air atmosphere.The optimum process parameters were further confirmed by in-situ monitoring of the temperature during the whole reaction process.The mixtures with Cu,Fe and Ni as the main components were obtained after the treatment of WMMs in the molten alkali.These mixed metals were further separated into copper-rich and ferronickel-rich metals by physical magnetic separation.Moreover,the precious metals Au and Ag were enriched in Cu-rich alloys.This work provided an efficient and environment-friendly method for metal recycling from WMMs.
基金supported by the National Natural Science Foundation of China(62203431)。
文摘The widespread usage of Cyber Physical Systems(CPSs)generates a vast volume of time series data,and precisely determining anomalies in the data is critical for practical production.Autoencoder is the mainstream method for time series anomaly detection,and the anomaly is judged by reconstruction error.However,due to the strong generalization ability of neural networks,some abnormal samples close to normal samples may be judged as normal,which fails to detect the abnormality.In addition,the dataset rarely provides sufficient anomaly labels.This research proposes an unsupervised anomaly detection approach based on adversarial memory autoencoders for multivariate time series to solve the above problem.Firstly,an encoder encodes the input data into low-dimensional space to acquire a feature vector.Then,a memory module is used to learn the feature vector’s prototype patterns and update the feature vectors.The updating process allows partial forgetting of information to prevent model overgeneralization.After that,two decoders reconstruct the input data.Finally,this research uses the Peak Over Threshold(POT)method to calculate the threshold to determine anomalous samples from normal samples.This research uses a two-stage adversarial training strategy during model training to enlarge the gap between the reconstruction error of normal and abnormal samples.The proposed method achieves significant anomaly detection results on synthetic and real datasets from power systems,water treatment plants,and computer clusters.The F1 score reached an average of 0.9196 on the five datasets,which is 0.0769 higher than the best baseline method.
文摘In order to implement semantic mapping of database metasearch engines, a system is proposed, which uses ontology as the organization form of information and records the new words not appearing in the ontology. When the new word＇ s frequency of use exceeds the threshold, it is added into the ontology. Ontology expansion is implemented in this way. The search process supports ＂and＂ and ＂or＂ Boolean operations accordingly. In order to improve the mapping speed of the system, a memory module is added which can memorize the recent query information of users and automatically learn the user＇ s query interest during the mapping which can dynamically decide the search order of instances tables. Experiments prove that these measures can obviously reduce the average mapping time.
基金This work is supported in part by the National Key Research and Development Program under Grant No. 2016YFB1000104the National Natural Science Foundation of China under Grant No. 61672345the HighTech Support Program from Shanghai Committee of Science and Technology under Grant No. 19511121100.
文摘Persistent indexing structures are proposed in response to emerging non-volatile memory(NVM)to provide high performance yet durable indexes.However,due to the lack of real NVM hardware,many prior persistent indexing structures were evaluated via emulation,which varies a lot across different setups and differs from the real deployment.Recently,Intel has released its Optane DC Persistent Memory Module(PMM),which is the first production-ready NVM.In this paper,we revisit popular persistent indexing structures on PMM and conduct comprehensive evaluations to study the performance differences among persistent indexing structures,including persistent hash tables and persistent trees.According to the evaluation results,we find that Cacheline-Conscious Extendible Hashing(CCEH)achieves the best performance among all evaluated persistent hash tables,and Failure-Atomic ShifT B+-Tree(FAST)and Write Optimal Radix Tree(WORT)perform better than other trees.Besides,we find that the insertion performance of hash tables is heavily influenced by data locality,while the insertion latency of trees is dominated by the flush instructions.We also uncover that no existing emulation methods accurately simulate PMM for all the studied data structures.Finally,we provide three suggestions on how to fully utilize PMM for better performance,including using clflushopt/clwb with sfence instead of clflush,flushing continuous data in a batch,and avoiding data access immediately after it is flushed to PMM.
