Skip to main content


         This documentation site is for previous versions. Visit our new documentation site for current releases.      
 

Training an email parser

Updated on May 17, 2024

Add training data to the pxEmailParser model to train the model with examples from your domain. As a result, the model is better at recognizing the signature, email body, and disclaimer in emails that are typical for your business.

  1. In Dev Studio, search for pxEmailParser, and then open the model.
  2. On the pxEmailParser tab, click Save as.
  3. In the Save As Decision Data window, select an open ruleset, and choose the appropriate label and identifier.
  4. Click Open in Prediction Studio.
  5. For the relevant language model, click DownloadTraining data (zip).
  6. Open the training file in Excel, and then add rows in the same format as the other rows.
    <START:entityType> denotes the start of an entity. <END> denotes the end of an entity. This model uses paragraph extraction, so ensure that each <START:entityType> is at the beginning of a line, and each <END> tag is at the end of a line. Ensure that each tag has a space before and after.
  7. Remove any rows if required.
    If you believe that the model works better with only domain-specific data, then you can choose to remove the default data that the model comes with.
  8. In Prediction Studio, click Models, and then open the relevant model.
  9. Click Create with machine learning.
  10. In the Source selection step, click Upload data source, and then upload the file.
  11. Complete the model update with the necessary model parameters.
For example: Here are some examples of how an email is parsed:
EmailExtracted text

Hi,

Can we have a call tomorrow to discuss on this recovery data set.

Thanks & Regards

Dave

-----------------------------------------------------

This message has been prepared by a Sales or Trading function of one or more affiliates of the Bank and is not the product of the Research Dept. It is not a research report.

This should not be construed as an offer to sell or the solicitation of an offer to buy any security in any jurisdiction where such an offer or solicitation would be illegal. It does not constitute a recommendation or take into account the investment objectives, financial conditions, or needs of individual clients.

This email and any files transmitted with it are confidential and intended solely for the person or entity to whom they are addressed and may contain confidential and/or privileged material.

Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon this information by persons or entities other than the intended recipient is prohibited. If you have received this email in error please contact the sender and delete the material from any computer.

<START:body> Hi,

Can we have a call tomorrow to discuss on this recovery data set. <END>

<START:signature> Thanks & Regards

Dave <END>

<START:disclaimer>

-----------------------------------------------------

This message has been prepared by a Sales or Trading function of one or more affiliates of the Bank and is not the product of the Research Dept. It is not a research report.

This should not be construed as an offer to sell or the solicitation of an offer to buy any security in any jurisdiction where such an offer or solicitation would be illegal. It does not constitute a recommendation or take into account the investment objectives, financial conditions, or needs of individual clients.

This email and any files transmitted with it are confidential and intended solely for the person or entity to whom they are addressed and may contain confidential and/or privileged material.

Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon this information by persons or entities other than the intended recipient is prohibited. If you have received this email in error please contact the sender and delete the material from any computer. <END>

Hi John,

We could replicate it in our environment.

Could you please provide the environment details?

Thanks

Mark

<START:body> Hi John,

We could replicate it in our environment.

Could you please provide the environment details? <END>

<START:signature> Thanks

Mark <END>

Hello,

The last merge into Cambridge was performed Friday.

Please change your bug status accordingly.

DISCLAIMER:

The information contained in this e-mail message is for the use of the addressee and is solely intended for the person to whom it has been sent. This message may contain legally privileged and confidential information which may not be made public. If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the same. Internet e-mails are not necessarily secure. ADP does not accept responsibility for changes made to this message after it was sent. ADP may monitor e-mails for business and operational purposes.

<START:body> Hello,

The last merge into Cambridge was performed Friday.

Please change your bug status accordingly. <END>

<START:disclaimer>

DISCLAIMER:

The information contained in this e-mail message is for the use of the addressee and is solely intended for the person to whom it has been sent. This message may contain legally privileged and confidential information which may not be made public. If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the same. Internet e-mails are not necessarily secure. ADP does not accept responsibility for changes made to this message after it was sent. ADP may monitor e-mails for business and operational purposes. <END>

What to do next: Test the email parser to see if the model processes text as expected. For more information, see Testing an email parser.

Have a question? Get answers now.

Visit the Support Center to ask questions, engage in discussions, share ideas, and help others.

Did you find this content helpful?

Want to help us improve this content?

We'd prefer it if you saw us at our best.

Pega.com is not optimized for Internet Explorer. For the optimal experience, please use:

Close Deprecation Notice
Contact us