For the intelligent dialogue system of customer service robots in power supply business halls, a large-scale dataset of power business user intents was constructed. The dataset includes 9 577 user queries and their labeling categories. First, the real voice data collected from the power supply business halls were cleaned, processed and filtered. In order to enable the data to drive the study of deep learning models related to intent classification, the data were labeled and augmented with high quality by the professionals according to the background knowledge of power business. In the labeling process, 35 types of service category labels were defined according to power business. In order to test the practicability and effectiveness of the proposed dataset, several classical models of intent classification were used for experiments, and the obtained intent classification models were put in the dialogue system. The classical Text classification model-Recurrent Convolutional Neural Network (Text-RCNN) was able to achieve 87.1% accuracy on this dataset. Experimental results show that the proposed dataset can effectively drive the research on power business related dialogue systems and improve user satisfaction.

Single-scale edge detection operator itself is sensitive to noise, which leads to little difference between the real and false edge, so the edge detected by it is not accurate, because ground object character is complex and thin ground object is intermingled with noise in real environment. Therefore, a new multi-scale fused edge detection algorithm based on conflict redistribution DSmT was proposed in this paper. First, multi-scale edge measure was extracted and then evidence theory was brought in. The basic belief assignment of multi-scale edge measure was constructed by a new method of bidirectional exponent and then fused by conflict redistribution DSmT combination rule. At last, edge points were extracted by multiple thresholds. The simulation with both optical and Synthetic Aperture Radar （SAR） images shows that the edge detection method of this paper suppresses noise effectively, while preserving rich details.