README.md

Question Understanding (QU)

Module to create an intent-explicit form of the current question and the corresponding conversational history.

• Create your own QU module
  • inference_on_turn function
  • inference_on_conversation function
  • train function

Create your own QU module

You can inherit from the QuestionUnderstanding class and create your own QU module. Implementing the functions inference_on_turn and inference_on_conversation is sufficient for the pipeline to run properly. You might want to implement your own training procedure for your module via the train function though.

Further, you need to instantiate a logger in the class, which will be used in the parent class. Alternatively, you can call the init method of the parent class.
Also, make sure to use the use_gold_answers parameter properly in your derived class. This parameter will be given as parameter when initializing the module.

inference_on_turn function

Inputs:

• turn: the current turn for which the intent-explicit representation should be generated.
• history_turns: the previous turns in the conversation. Can be used to generate the intent-explicit form. List of turn dictionaries.

Description:
This method is supposed to generate an intent-explicit form of the current question, given the conversational history. Please make sure that the class parameter use_gold_answers controls whether the gold answer(s) (in turn["answers"]) or the predicted answer(s) (in turn["pred_answers"]) are used.

Output:
Returns the turn. Make sure to store the intent-explicit representation of the information need in turn["structured_representation"].

inference_on_conversation function

Inputs:

• conversation: the conversation for which the intent-explicit representation should be generated.

Description:
This method is supposed to generate intent-explicit forms for all turns in the conversation. In the method, you can keep track of the conversational history (e.g. using a list). Please make sure that the class parameter use_gold_answers controls whether the gold answer(s) (in turn["answers"]) or the predicted answer(s) (in turn["pred_answers"]) are used.

Output:
Returns the conversation. Make sure to store the intent-explicit representation of the information need for every turn in the conversation, in turn["structured_representation"].

[Optional] train function

Inputs: NONE

Description:
If required, you can train your QU module here. You can make use of whatever parameters are stored in your .yml file.

Output: NONE