- Fill content featurs in X.json , educativeness in y.json and manually annotated features in X_manual.json. This can be done using FillContentFeatures.py
- Run the different tasks of the experiment by using fuctions provided in SVM_Tests.py. A sample code for running is provided in RunExperiment.py.
Accuracy for task_1: 0.57517314062 +- 0.0650688058016
Accuracy for task_2: 0.752378801566 +- 0.0661595095201
Accuracy for task_3 (Relevance): 0.573833182776 +- 0.0504209371291
Accuracy for task_3 (Classification_Example): 0.754953327311 +- 0.04273631597
Accuracy for task_3 (Classification_Definition): 0.791839807287 +- 0.0553806585448
Accuracy for task_3 (Classification_Illustrations): 0.772131887986 +- 0.0623932603013
Accuracy for task_3 (Classification_QA): 0.91251129178 +- 0.0250967782722
Accuracy for task_3 (Classification_Other): 0.775865703101 +- 0.0415203537873
Accuracy for task_3 (Source_ClassWebpage): 0.864664257754 +- 0.0337293821065
Accuracy for task_3 (Source_Encyclopedia): 0.953131586871 +- 0.0246161531611
Accuracy for task_3 (Source_Blog): 0.906338452273 +- 0.0283847339988
Accuracy for task_3 (Source_Forums): 0.917464619091 +- 0.020026443568
Accuracy for task_3 (Source_Book): 0.862029509184 +- 0.0281566017819
Accuracy for task_3 (Source_Presentation): 0.951987353207 +- 0.0363034536084
Accuracy for task_3 (Source_Publication): 0.884251731406 +- 0.0338900433139
Accuracy for task_3 (Source_HowTo): 0.780879253237 +- 0.0416820806138
Accuracy for task_3 (Source_Manual): 0.863249021379 +- 0.031147053989
Accuracy for task_3 (Source_Other): 0.76475459199 +- 0.0453978473945
Accuracy for task_4: 0.582535380909 +- 0.0685755202386