Data Classification Parameters

Data Classification Parameters

Parameters for |Data Classification Evaluation

The parameters for the evaluation of sentiment analysis include various terms. The terms are True positives, true negatives, false negatives and false positives.These are the terms that are used to compare the class labels assigned to documents with the classes the items actually belong to by a classifier.True positive terms are truly classified as positive terms.False positive are not labeled by the classifier as a positive class but should have been.True negative terms are correctly labeled as in a negative class by the classifier.False negative terms are those terms that are not labeled by the classifier as belonging to negative class but should have been classified.Confusion Matrix contains these terms that are used for evaluation.

 

Correct Labels

 Positive  Negative
 

Classified Labels

Positive True positive False positive
Negative False negative True negative

 

Fig. Contingency table

Following are the parameters for evaluation of performance:

Precision and recall 

Precision and recall are the two metrics that are widely for evaluating performance in text mining, and in text analysis field like information retrieval. These parameters are used for measuring exactness and completeness respectively.

F-measure

F-Measure is the harmonic mean of precision and recall. The value calculated using F-measure is a balance between precision and recall.

Accuracy

Accuracy is the common measure for classification performance. Accuracy can be measured as correctly classified instances to the total number of instances, while error rate uses incorrectly classified instances instead of correctly classified instances.

 

 

 

 

 

Share

Leave a Reply

Your email address will not be published. Required fields are marked *