Detecting Email Commitments using Natural Language Processing
Aug 2012 - Dec 2012
Design and development of an intelligent system that identifies the business tasks and commitments in the e-mail conversations using machine learning algorithms and natural language processing.I'm a paragraph. Click here to add your own text and edit me. I’m a great place for you to tell a story and let your users know a little more about you
TEAM
Anup Kalia
Siddhant Shah
Sujay Sarkhel
MY ROLES
Researcher
Tags Annotator
DELIVERABLES
Final Report
Presentatio
This semester-long project was developed for the Natural language dailog systems class. Business interactions in most organizations have transitioned from traditional pen and paper usage to much more sophisticated, environment-friendly email conversations. Such formal interactions, in business parlance, often involve defining, assigning or reporting action items. Many such action items dont signal the creation, delegation, discharge or even cancellation of commitments amongst the participants. There comes and increasing necessity to manage such e-mail conversations and the action items.
Approach
We built an intelligent agent that automatically identifies relevant action items and determines commitments within such email conversations.
Our model is built using natural language processing (NLP) and machine learning (ML) techniques. We identified tasks, task owners, and deadlines from amessage using NLP. We subsequently labeled sentences as indicating the creation,delegation, cancellation of commitment or discharge of a commitment.
Finally, to build the training dataset, we used several classifiers such as NaïveBayesian (NB), Logistic Regression (LR), and Support Vector Machine (SVM) classifier. For our evaluation, we used the 10-Fold cross validation technique.