arrow_back BoF on Interpretability of ML Models
Birds of Feather (BOF) session: Intent classification and personalization
Submitted by Ramanan Balakrishnan (@ramananbalakrishnan) via Abhishek Balaji (@booleanbalaji) on Monday, 15 July 2019
Session type: Birds of a Feather session of 1 hour Session type: BOF session of 1 hour
When it comes to developing a comprehensive natural language understanding system, intent classification is one of the first challenges to overcome. Without developing an understanding of the context of a text, it becomes almost impossible to interpret entities that may be recognized in later stages.
One of the main reasons intent classification is popular is also its use in achieving personalization for each user. By right suggestions at the right time, a model can work wonders on improving the user experience.
The intent (pun intended) of this BoF is to discuss user stories in developing, deploying and maintaining such detection and personalization systems.
- Applications that may require/benefit from intent classification/personalization methods
- Bias in ML
- Bias in intent classification.
- Fairness in ML.
- Production war stories
- Questions to Ask:
- What was the first model you pushed out.
- A general overview of the field of intent classification and personalization as a bare minimum.
- Which applications commonly benefit from these approaches, when to use them, and of-course, when not to use them.
- Production war stories would also be shared as learning experiences.
Who should attend?
- People with background/interest in personalised applications and NLP, or are new to the world of personalisation/intent classification
- People building chatbots and/or search engines, analysts working with customer reviews or user sentiments, and working on building recommendation engines.
- Background in data and ML.
- Some experience using these topics across any domain - with some level of real-world experience
- Aditya Patel
- Ishita Mathur
- Ramanan Balakrishnan
- Maulik Soneji