Research Article

Text Classification Using Novel Term Weighting Scheme-Based Improved TF-IDF for Internet Media Reports

Table 12

Performance on the Fudan corpus using the SVM classifier.

SVMPrecisionF1 scoreRecall
TF-IDFTF-IADFTF-IADFnormTF-IADF+TF-IADF+normTF-IDFTF-IADFTF-IADFnormTF-IADF+TF-IADF+normTF-IDFTF-IADFTF-IADFnormTF-IADF+TF-IADF+norm

C1196.1095.9896.1396.24%96.24%92.0692.9992.8491.75%91.75%94.0394.4694.4593.94%93.94%
C15100.00100.00100.00100.00%100.00%9.099.099.099.09%9.09%16.6716.6716.6716.67%16.67%
C16100.00100.00100.00100.00%100.00%3.573.573.573.57%3.57%6.906.906.906.90%6.90%
C17100.00100.00100.00100.00%100.00%18.5222.2222.2218.52%18.52%31.2536.3636.3631.25%31.25%
C1995.6495.9395.5395.50%95.56%98.5398.9799.1298.53%98.31%97.0697.4397.2996.99%96.92%
C23100.00100.00100.00100.00%100.00%5.885.885.885.88%5.88%11.1111.1111.1111.11%11.11%
C29100.0096.1595.65100.00%100.00%8.4842.3737.295.09%3.39%15.6358.8253.669.68%6.56%
C387.1088.8589.1787.18%86.90%95.5595.5595.4295.28%94.74%91.1392.0892.1991.05%90.65%
C3195.6196.9797.7795.80%95.86%96.6397.2196.9697.29%97.04%96.1297.0997.3696.54%96.45%
C3294.1794.1093.4794.43%94.37%96.3896.6796.5896.18%95.21%95.2695.3795.0095.30%94.79%
C3490.2090.1890.2389.67%87.03%94.8895.1994.6394.88%94.75%92.4892.6292.3892.20%90.73%
C35100.0095.2492.31100.00%100.00%23.0838.4646.1511.54%7.69%37.5054.8061.5420.69%14.29%
C36100.00100.00100.00100.00%100.00%1.8913.2115.091.89%1.89%3.7023.3326.233.70%3.70%
C37100.00100.00100.00100.00%100.00%5.265.265.265.26%5.26%10.0010.0010.0010.00%10.00%
C3882.6582.3682.7883.19%82.66%95.6194.6494.6495.52%95.22%88.6688.0788.3188.93%88.50%
C3985.3688.6987.6884.86%85.93%96.7396.9796.4996.97%96.89%90.6992.6591.8890.51%91.08%
C4100.00100.00100.00100.00%100.00%5.885.885.885.88%2.94%11.1111.1111.1111.11%5.71%
C566.6766.6766.6766.67%66.67%3.283.283.283.28%3.28%6.256.256.256.25%6.25%
C688.8988.8990.0088.89%88.89%17.7817.7820.0017.78%17.78%29.6329.6332.7329.63%29.63%
C782.7881.3080.9982.43%81.70%68.8072.4473.7268.16%65.81%75.1576.6177.1874.62%72.90%