多视角网页分类数据集构建及性能评估

孙辰星, 刘伟, 卢彬, 梁诗宇, 诸云强, 甘小莺

Multi⁃view webpage classification dataset construction and evaluation

Chenxing Sun, Wei Liu, Bin Lu, Shiyu Liang, Yunqiang Zhu, Xiaoying Gan

表5 Fusion模型对分布不均衡的网页主题进行分类的准确率比较

Table 5 Classification accuracy of different topics with category imbalance by Fusion model

类别 下采样 样本生成 损失重加权 平均 75.56% 79.26% 82.90% 地质学 78.36% 82.24% 84.17% 地球物理学 77.35% 81.92% 83.67% 地理学 68.71% 74.35% 78.17% 地质资源学 68.95% 74.21% 77.70%