ARTFEED — Contemporary Art Intelligence

TeCQR: A Conversational Model for Related Question Retrieval in cQA

other · 2026-04-29

Researchers propose TeCQR, a model for related question retrieval in community question answering (cQA) platforms like Stack Overflow. Unlike traditional static approaches, TeCQR leverages conversations to capture fine-grained question representations. It builds conversations using tag-enhanced clarifying questions and includes a noise tolerance model to evaluate semantic similarity between questions and tags, handling noisy data effectively.

Key facts

  • TeCQR is a related question retrieval model using conversations.
  • It targets cQA platforms like Stack Overflow.
  • Traditional approaches are static and neglect interaction.
  • Conversations help distinguish fine-grained question representations.
  • Tag-enhanced clarifying questions (CQs) are used to build conversations.
  • A noise tolerance model evaluates semantic similarity between questions and tags.
  • The model handles noisy data effectively.

Entities

Sources