Search

Sequence labeling is the basis for many tasks in natural language processing (NLP). It plays an important role in tasks such as word segmentation, named entity recognition (NER), and part-of-speech (POS)tagging. The current mainstream method for sequence labeling is to combine neural network with conditional random field (CRF). The common model is usually a bidirectional RNN-CRF model, which can solve the problem that the labeling task with traditional method cannot be combined well with the context. This paper proposes a Chinese sequence labeling model based on bidirectional GRU-CNN-CRF, which can pay more attention to local features and context relationships, and has better performance in word segmentation and NER. This paper takes the corpus provided by Chinese Wikipedia as the training data set and preprocesses the text by word embedding. The data are then processed through a three-tier architecture of bidirectional Gated Recurrent Unit (GRU), Convolution Neural Network (CNN)and CRF, and finally complete the task of sequence annotation. Compared with the traditional Chinese word segmentation system, this method is more accurate. And it performs better than bidirectional GRU-CRF model on NER issues.

In recent decades, the rapid development of neuroimaging technology has provided an effective and noninvasive method for exploring the structure and functional connectivity of the human brain. Corresponding neural image data mining plays an increasingly important role in predicting neurological and psychiatric diseases because it may find clinically significant structural patterns that can promote the understanding and diagnosis of neurological and psychiatric diseases. Most of the recent researches focus on sub-graph models of brain networks. However, due to the high complexity of brain network connections, the number of sub-graphs increases exponentially with the increase of graph patterns. Traditional sub-graph mining is both time-consuming and memory-intensive. In this paper, by measuring the difference in brain network connectivity between normal individuals and patients, we propose a coritivity-based brain network mining method to help diagnose neurological or psychiatric diseases. Since solving the network coritivity is an NP-Hard problem, we propose a heuristic algorithm based on the Max-Min Ant System(MMAS), which can quickly give the approximate coritivity of the network. We conducted experiments on three real-world brain network datasets for disease diagnosis. The experiment results show that our method achieves great results and can provide meaningful suggestions.

The authors combine feature-based kernel with extension path graph kernel into a multiple kernels learning method. Feature-based kernel method, extension path graph kernel method and multiple kernels learning method are conducted on experiment on the most authoritative five evaluation corpuses. Experimental results indicate that the performance of the fused kernel method in five corpus sets is superior to that of the two separate single-kernel method.

Xi'an Qujiang New District is a new district with cultural and tourism industries as its main developing engines. In recent years, the district has grown into one of the major landmark regions of western China based on its considerable advantage in cultural relics, culture and tourism. In order to discuss the develop mode of Qujiang New District and serve China's urban development, this research focusing on Qujiang New District landscape ecology and using the established landscape hierarchical evaluation index system, conducted a survey on people's satisfaction and suggestions to Qujiang environment. The result shows that people are satisfied with the overall planning, overall regional eco-environment, and completeness of life and service facilities. However, people are less satisfied with organizational management, water quality environment, and food culture diversity. It is suggested that construction of the Qujiang New District should be carried out more comprehensively in theme. More green energy should be used; quality and appearance of construction should be improved. Harmonize residential environment with the colors, architecture type and environment landscape of surrounding historical buildings. Increase investment in public resource, make the management more people-oriented and improve management efficiency. This survey provides evaluation index and scientific basis to improve human-environment relation in cities.

As urban construction development, urban landscape ecological environment is drawing wide attention. Therefore, it is important to build urban landscape ecological evaluation system adapt to city development. This paper comprehensively studied major problems in urban landscape construction. By giving overall consideration to regional ecological environment, social cultural and residential environment, man-made physical environment, eco-civilization construction as well as organizational management, a landscape hierarchical evaluation index system composed of one target level, four standard levels and 27 indexes is built. Based on this index system and through comprehensive evaluation, we can provide basic statistics for landscape ecological planning in city planning and construction as well as evaluation index and scientific basis in order to improve human-environment relation in cities.