Journals
  Publication Years
  Keywords
Search within results Open Search
Please wait a minute...
For Selected: Toggle Thumbnails
English Books Automatic Classification According to CLC
JIANG Yanting
Acta Scientiarum Naturalium Universitatis Pekinensis    2023, 59 (1): 11-20.   DOI: 10.13209/j.0479-8023.2022.070
Abstract347)   HTML    PDF(pc) (797KB)(175)       Save
Faced with lacking of English books annotated with CLC (Chinese Library Classification) label and imbalance data, this paper combines augmentation strategies from library, information and general fields: 1) classification mapping from Library of Congress Classification (LCC) to CLC; 2) semantic enhancement based on Chinese-English parallel thesaurus; 3) punctuation or 4) conjunction inserting to initial texts. Experiments show that combining 4 strategies can optimize the performance of models on test set. Accuracy and Macro-F1 respectively increase by 3.61 and 3.35 percentage points. Comprehensive methods is superior to other text enhancement strategies. By BERT word embeddings visualization and words information entropy computing, this paper inferred that the reason why punctuation or conjunction inserting works was the various adjacent words and connection function in grammar.
Related Articles | Metrics | Comments0
Characteristics of Pollutants Flush on Different Types of Underlying Surface in Changzhou
ZHANG Xiangli, ZHAO Zhijie, QIN Huapeng, JIANG Yan
Acta Scientiarum Naturalium Universitatis Pekinensis    2018, 54 (3): 644-654.   DOI: 10.13209/j.0479-8023.2017.176
Abstract1159)   HTML9)    PDF(pc) (1248KB)(301)       Save

The purpose of the study is flush characteristics of urban runoff pollutant on different underlying surface. Select 5 typical hardened surfaces in Changzhou City and monitor the change process of rainfall runoff pollutant from March to August in 2015. The results showed that event mean concentrations (EMC) of pollutants in road runoff were higher than that of roof runoff. For road runoff, SS concentration was higher than the water quality standards by 1.34 times; COD concentration was higher than the water quality standards by 2.59 times. For the roof runoff, COD concentration was higher than the water quality standards by 1.8 times; and TN concentration was higher than the water quality standards by 2.6 times. For the roof runoff, the dissolved-bound fraction was 72.78% for COD, 57.99% for TN. For road runoff, the dissolved-bound fraction was 61.59% for TN. The pollutant concentrations were commonly higher at the initial stage, while decreased with prolonging of the rainfall time and gradually became stable at the later stage. The initial concentrations of pollutants from the underlying surface were as follows: concrete ground, asphalt ground, paved ground, flat roof and slope roof. During the rain flush, the concentration of pollutants on the underlying surface increases with the increase of the intensity of the rain which was fluctuated. The intensity of the first flush intensity varied by surface and was most intense for the flat roof, followed by the slope roof, then the asphalt road and finally the concrete road. Rainfall in pre-period was intensive and pollutant concentration was exponentially attenuated. When rainfall changes smoothly, the concentration of pollutants was stable firstly and then attenuation. When the rainfall was sparse in pre-period and intensive in the late-period, the curves of pollutant change type was multi-peak type. The index flush model had good effect to pollutant runoff simulation, the flush coefficients of COD on the slope roof, flat roof, and concrete road were 0.871, 0.765, and 0.025 mm−1, the roof flush intensity was much larger than the ground. The flush coefficient of dissolved-bound of COD and granular-bound of COD was similar on the slope roof. The flush coefficient of granular-bound of COD was greater than the dissolved-bound of COD on the flat roof and the concrete road.

Related Articles | Metrics | Comments0