Discover published sets by community

Discourse Analysis involves understanding language beyond the sentence level, looking at the interplay of utterances in a conversation to interpret meaning, coherence, and structure.

Response Generation is the task of producing an appropriate reply to the user's input. This can vary from selecting a canned response to generating a new sentence using natural language generation techniques.

Multi-modal Dialogue Systems can process and generate multiple types of data, such as text, audio, images, and video, to interact with users. An example is a system that handles both spoken input and touch input.

Slot Filling involves collecting pieces of information required to fulfill a request. For example, when booking a hotel, slots for check-in date, check-out date, and number of guests would be filled.

Dialogue State Tracking keeps track of the user's goals and requests throughout the conversation to understand the context of the dialogue. It updates the state as the conversation progresses.

Entity Extraction refers to the process of identifying and classifying key elements from user input, such as dates, names, locations, etc. For instance, extracting 'New York City' as a location entity from a travel query.

Speech Recognition in dialogue systems refers to the ability of a system to convert spoken language into text. It is an essential first step for voice-driven systems to understand user input.

Dialogue Management is the component that controls the flow of the conversation in a dialogue system, ensuring the system responds coherently. It takes into account the dialogue history and current state.

Intent Recognition involves determining the purpose behind a user's input in a dialogue system. For example, detecting if a user's intent is to book a flight or to ask for weather information.