《Table 5 Results on Finance-Test》
本系列图表出处文件名:随高清版一同展现
《A document-level model for tweet event detection》
To evaluate the robustness of the method,the results(forth line)are compared with the baseline methods of Ref.[9]on the sports dataset,shown in Table 6.Sensor(third line)shows the proposed method in Ref.[9],which is discussed in Section 3.The first two lines in Table 6 show results of two strong baseline methods,which are selected from Ref.[9].The LDA method is a topic model based on event detection method,which uses LDA(Latent Dirichlet Allocation)for clustering.Doc-p is a document-based topic detection method by Locality Sensitive Hashing.The method improves the topic recall from 76.92%by Sensor and Doc-p to 88.89%.Keyword precision and recall are also improved compared to all baseline methods.The improvement verifies the effectiveness of our method on the sports domain,which shows the robustness of the method.
图表编号 | XD0020600300 严禁用于非法目的 |
---|---|
绘制时间 | 2018.06.01 |
作者 | 秦彦霞、Zhang Yue、Zhang Min、Zheng Dequan |
绘制单位 | School of Computer Science and Technology,Harbin Institute of Technology、Information Systems Technology and Design,Singapore University of Technology and Design、School of Computer Science and Technology,Soochow University、School of Computer Science and Te |
更多格式 | 高清、无水印(增值服务) |