Jiao Long
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jiao Long has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Dongmei Li
School of Information Science and Technology, Beijing Forestry University, Beijing 100083, China
1 shared publication - Jianxin Wang
School of Information, Beijing Forestry University, Beijing 100083, China, China
1 shared publication - Jintao Qu
School of Information Science and Technology, Beijing Forestry University, Beijing 100083, China
1 shared publication - Xiaoping Zhang
National Data Center of Traditional Chinese Medicine China Academy of Chinese Medical Sciences Beijing China
1 shared publication - Yanan Gao
School of Computer Science and Technology, Beijing Institute of Technology, Beijing, China
1 shared publication
Latest Publications
- Open Access
- 23 May 2022
- Published by Wiley in Evidence-Based Complementary and Alternative Medicine
- Vol. 2022, 1-9
Chinese Clinical Named Entity Recognition with ALBERT and MHA Mechanism
Traditional clinical named entity recognition methods fail to balance the effectiveness of feature extraction of unstructured text and the complexity of neural network models. We propose a model based on ALBERT and a multihead attention (MHA) mechanism to solve this problem. Structurally, the model first obtains character-level word embeddings through the ALBERT pretraining language model, then inputs the word embeddings into the iterated dilated convolutional neural network model to quickly extract global semantic information, and decodes the predicted labels through conditional random fields to obtain the optimal label sequence. Also, we apply the MHA mechanism to capture intercharacter dependencies from multiple aspects. Furthermore, we use the RAdam optimizer to boost the convergence speed and improve the generalization ability of our model. Experimental results show that our model achieves an F1 score of 85.63% on the CCKS-2019 dataset—an increase of 4.36% compared to the baseline model.
- 23 May 2022
- Published by Wiley in Evidence-Based Complementary and Alternative Medicine
- Vol. 2022, 1-9
Open AccessChinese Clinical Named Entity Recognition with ALBERT and MHA Mechanism
Traditional clinical named entity recognition methods fail to balance the effectiveness of feature extraction of unstructured text and the complexity of neural network models. We propose a model based on ALBERT and a multihead attention (MHA) mechanism to solve this problem. Structurally, the model first obtains character-level word embeddings through the ALBERT pretraining language model, then inputs the word embeddings into the iterated dilated convolutional neural network model to quickly extract global semantic information, and decodes the predicted labels through conditional random fields to obtain the optimal label sequence. Also, we apply the MHA mechanism to capture intercharacter dependencies from multiple aspects. Furthermore, we use the RAdam optimizer to boost the convergence speed and improve the generalization ability of our model. Experimental results show that our model achieves an F1 score of 85.63% on the CCKS-2019 dataset—an increase of 4.36% compared to the baseline model.
- 1 August 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 48, 2673-2680
The Optimal Layout of Fireproof Sensor Network Based on Improved Cuckoo Algorithm on Mountainous Forest Land
The fireproof sensor network plays a more and more important role on monitoring and detecting forest fire. Different from previous researches which focus mostly on plane or single slope, this work deals with the scenarios of unsmooth moutainous land, which are the majority of application situations in forest fire prevention. The layout of wireless sensor nodes in the complex forest slope is a typical NP-complete problem. Another difficulty is that varied terrain factors bring huge obstacles for calculating the fitness value of optimization algorithm. To overcome these difficulties, firstly, this paper provides a computing method of fitness value, based on fire-response time on the simulated forest land. Secondly, we analyze the defects on parameter adjustment of original NMCSA (Novel Modified Cuckoo Search Algorithm), and improve it in two aspects: (1) cutting down the step-size factor errors by using mean operator and (2) redefining the updating method of discarding probability. The simulated experiments indicate that the improved NMCSA outperforms the previous algorithms in convergence speed and accuracy. What's more, the averaged response time of the generated layout by the new method is shorter.
- 1 August 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 48, 2673-2680
The Optimal Layout of Fireproof Sensor Network Based on Improved Cuckoo Algorithm on Mountainous Forest Land
The fireproof sensor network plays a more and more important role on monitoring and detecting forest fire. Different from previous researches which focus mostly on plane or single slope, this work deals with the scenarios of unsmooth moutainous land, which are the majority of application situations in forest fire prevention. The layout of wireless sensor nodes in the complex forest slope is a typical NP-complete problem. Another difficulty is that varied terrain factors bring huge obstacles for calculating the fitness value of optimization algorithm. To overcome these difficulties, firstly, this paper provides a computing method of fitness value, based on fire-response time on the simulated forest land. Secondly, we analyze the defects on parameter adjustment of original NMCSA (Novel Modified Cuckoo Search Algorithm), and improve it in two aspects: (1) cutting down the step-size factor errors by using mean operator and (2) redefining the updating method of discarding probability. The simulated experiments indicate that the improved NMCSA outperforms the previous algorithms in convergence speed and accuracy. What's more, the averaged response time of the generated layout by the new method is shorter.