Search

Recently, researchers have extensively explored various methods for electronic medical record named entity recognition, including character-based, word-based, and hybrid methods. Nonetheless, these methods frequently disregard the semantic context of entities within electronic medical records, leading to the creation of subpar-quality clinical knowledge bases and obstructing the discovery of clinical knowledge. In response to these challenges, we propose a novel purpose-driven SoftLexicon-RoBERTa-BiLSTM-CRF (SLRBC) model for electronic medical records named entity recognition. SLRBC leverages the fusion of SoftLexicon and RoBERTa to incorporate the word lexicon information from electronic medical records into the character representations, enhancing the model’s semantic embedding representations. This purpose-driven approach helps achieve a more comprehensive representation and avoid common segmentation errors, consequently boosting the accuracy of entity recognition. Furthermore, we employ the classical BiLSTM-CRF framework to capture contextual information of entities more effectively. In order to assess the performance of SLRBC, a series of experiments on the public datasets of CCKS2018 and CCKS2019 were conducted. The experimental results demonstrate that SLRBC can efficiently extract entities from Chinese electronic medical records. The model attains F1 scores of 94.97% and 85.40% on CCKS2018 and CCKS2019, respectively, exhibiting outstanding performance in the extraction and utilization efficiency of clinical information.

Drones are widely used for wildlife monitoring. Deep learning algorithms are key to the success of monitoring wildlife with drones, although they face the problem of detecting small targets. To solve this problem, we have introduced the SE-YOLO model, which incorporates a channel self-attention mechanism into the advanced real-time object detection algorithm YOLOv7, enabling the model to perform effectively on small targets. However, there is another barrier; the lack of publicly available UAV wildlife aerial datasets hampers research on UAV wildlife monitoring algorithms. To fill this gap, we present a large-scale, multi-class, high-quality dataset called WAID (Wildlife Aerial Images from Drone), which contains 14,375 UAV aerial images from different environmental conditions, covering six wildlife species and multiple habitat types. We conducted a statistical analysis experiment, an algorithm detection comparison experiment, and a dataset generalization experiment. The statistical analysis experiment demonstrated the dataset characteristics both quantitatively and intuitively. The comparison and generalization experiments compared different types of advanced algorithms as well as the SE-YOLO method from the perspective of the practical application of UAVs for wildlife monitoring. The experimental results show that WAID is suitable for the study of wildlife monitoring algorithms for UAVs, and SE-YOLO is the most effective in this scenario, with a mAP of up to 0.983. This study brings new methods, data, and inspiration to the field of wildlife monitoring by UAVs.

Social media is important for providing text information; however, due to its informal and unstructured nature, traditional named entity recognition (NER) methods face the challenge of achieving high accuracy when dealing with social media data. This paper proposes a new method for social media named entity recognition with data augmentation. First, we pre-train the language model by using a bi-directional encoder representation of the transformer (BERT) to obtain a semantic vector of the word based on the contextual information of the word. Then, we obtain similar entities via data augmentation methods and perform substitution or semantic transformation on these entities. After that, the input into the Bi-LSTM model is trained and then fused and fine-tuned to obtain the best label. In addition, our use of the self-attentive layer captures the essential information of the features and reduces the reliance on external information. Experimental results on the WNUT16, WNUT17, and OntoNotes 5.0 datasets confirm the effectiveness of our proposed model.

Aspect-based sentiment analysis (ABSA) is a task in natural language processing (NLP) that involves predicting the sentiment polarity towards a specific aspect in text. Graph neural networks (GNNs) have been shown to be effective tools for sentiment analysis tasks, but current research often overlooks affective information in the text, leading to irrelevant information being learned for specific aspects. To address this issue, we propose a novel GNN model, MHAKE-GCN, which is based on the graph convolutional neural network (GCN) and multi-head attention (MHA). Our model incorporates external sentiment knowledge into the GCN and fully extracts semantic and syntactic information from a sentence using MHA. By adding weights to sentiment words associated with aspect words, our model can better learn sentiment expressions related to specific aspects. Our model was evaluated on four publicly benchmark datasets and compared against twelve other methods. The results of the experiments demonstrate the effectiveness of the proposed model for the task of aspect-based sentiment analysis.

Fundus blood vessel image segmentation plays an important role in the diagnosis and treatment of diseases and is the basis of computer-aided diagnosis. Feature information from the retinal blood vessel image is relatively complicated, and the existing algorithms are sometimes difficult to perform effective segmentation with. Aiming at the problems of low accuracy and low sensitivity of the existing segmentation methods, an improved U-shaped neural network (MRU-NET) segmentation method for retinal vessels was proposed. Firstly, the image enhancement algorithm and random segmentation method are used to solve the problems of low contrast and insufficient image data of the original image. Moreover, smaller image blocks after random segmentation are helpful to reduce the complexity of the U-shaped neural network model; secondly, the residual learning is introduced into the encoder and decoder to improve the efficiency of feature use and to reduce information loss, and a feature fusion module is introduced between the encoder and decoder to extract image features with different granularities; and finally, a feature balancing module is added to the skip connections to resolve the semantic gap between low-dimensional features in the encoder and high-dimensional features in decoder. Experimental results show that our method has better accuracy and sensitivity on the DRIVE and STARE datasets (accuracy (ACC) = 0.9611, sensitivity (SE) = 0.8613; STARE: ACC = 0.9662, SE = 0.7887) than some of the state-of-the-art methods.