The Regex classifier service helps in classifying certain patterns present in input text provided by the user and extracts out the pattern present in it along with corresponding labels. This classifier learns from your training data and then returns the labels which corresponds to patterns in text.
You can train a RegEx Classifier in less than 10 minutes.
For you to create a classifier that can work with your own data, you need to assemble and provide the training data in CSV format.
A column in the file represents an example record. First row of the record should be the class labels and rows below it represent its features.
The following table shows an example data from the sample CSV file:
firstname.lastname@example.org, 436 03 , 420-(440)144-7470, 3535700556764697
email@example.com, 622937, 7-(696)626-767, 5002352937442426
firstname.lastname@example.org, 89160-000, 55-(918)931-7571, 3566749105337259
For example, if your input text contains ‘email@example.com’ in between, then the regex classifier extracts out this text along with its class label ‘Email_id’. You can test the system with the sample training data provided in regex_sample.csv