Results.txt

Kim with tfidf + lsi
Accuracy
0.597
Precision
0.557666214383
Recall
0.842213114754
F1 Score
0.671020408163
Mean Squared Error
0.0899310240381
Explained variance
0.0830745537662
(Without lsi, the accuracy was terrible)

O'Mahony
Accuracy
0.57
Precision
0.543154761905
Recall
0.747950819672
F1 Score
0.629310344828
Mean Squared Error
0.103193149566
Explained variance
-0.117019278854


Liu
Accuracy
0.582
Precision
0.555379746835
Recall
0.719262295082
F1 Score
0.626785714286
Mean Squared Error
0.0944460892587
Explained variance
0.0221160835206



Zhang

Accuracy
0.496
Precision
0.489847715736
Recall
0.790983606557
F1 Score
0.605015673981
Mean Squared Error
0.105409123275
Explained variance
-0.0714352408212



tfidf unigrams with lemmatization but without lsi
Terrible accuracy

tfidf unigrams + lemmatization + lsi
Accuracy
0.549
Precision
0.520624303233
Recall
0.956967213115
F1 Score
0.674368231047
Mean Squared Error
0.0991283264928
Explained variance
0.00419150841297

sentiment
0 0.492033548109
1 0.507966451891
Accuracy
0.548
Precision
0.521126760563
Recall
0.909836065574
F1 Score
0.662686567164
Mean Squared Error
0.0986320458686
Explained variance
0.00655550821855


pos (basic pos: noun, verb, adj+adv, first person verbs, all 4 added)
Accuracy
0.568
Precision
0.541297935103
Recall
0.752049180328
F1 Score
0.629502572899
Mean Squared Error
0.0954461216704
Explained variance
0.0327418994006



Only combo with slightly higher accuracy than each by itself

Kim + Zhang
Accuracy
0.601
Precision
0.563120567376
Recall
0.813524590164
F1 Score
0.665549036044
Mean Squared Error
0.0982330156221
Explained variance
0.00916499580639