Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

CHOIR: A Chatbot-mediated Organizational Memory Leveraging Communication in University Research Labs

Sangwook Lee, Adnan Abbas, Yan Chen, Young-Ho Kim, Sang Won Lee

TL;DR

This work introduces CHOIR, a Slack-integrated chatbot that grounds questions and knowledge management in organizational documents to sustain lab memory. Through formative study and a one-month field deployment across four university labs, CHOIR demonstrates how document-grounded Q&A, knowledge extraction from conversations, Q&A sharing, and AI-assisted document updates can enhance retrieval and documentation while revealing privacy-awareness tensions. The findings highlight that AI-mediated OM can facilitate human communication and surface knowledge gaps, but privacy concerns and cultural factors shape how knowledge is shared and documented. The authors propose design implications for privacy-preserving awareness, contextualized knowledge capture, and AI mediation as a facilitator of collaboration across organizational boundaries.

Abstract

University research labs often rely on chat-based platforms for communication and project management, where valuable knowledge surfaces but is easily lost in message streams. Documentation can preserve knowledge, but it requires ongoing maintenance and is challenging to navigate. Drawing on formative interviews that revealed organizational memory challenges in labs, we designed CHOIR, an LLM-based chatbot that supports organizational memory through four key functions: document-grounded Q&A, Q&A sharing for follow-up discussion, knowledge extraction from conversations, and AI-assisted document updates. We deployed CHOIR in four research labs for one month (n=21), where the lab members asked 107 questions and lab directors updated documents 38 times in the organizational memory. Our findings reveal a privacy-awareness tension: questions were asked privately, limiting directors' visibility into documentation gaps. Students often avoided contribution due to challenges in generalizing personal experiences into universal documentation. We contribute design implications for privacy-preserving awareness and supporting context-specific knowledge documentation.

CHOIR: A Chatbot-mediated Organizational Memory Leveraging Communication in University Research Labs

TL;DR

This work introduces CHOIR, a Slack-integrated chatbot that grounds questions and knowledge management in organizational documents to sustain lab memory. Through formative study and a one-month field deployment across four university labs, CHOIR demonstrates how document-grounded Q&A, knowledge extraction from conversations, Q&A sharing, and AI-assisted document updates can enhance retrieval and documentation while revealing privacy-awareness tensions. The findings highlight that AI-mediated OM can facilitate human communication and surface knowledge gaps, but privacy concerns and cultural factors shape how knowledge is shared and documented. The authors propose design implications for privacy-preserving awareness, contextualized knowledge capture, and AI mediation as a facilitator of collaboration across organizational boundaries.

Abstract

University research labs often rely on chat-based platforms for communication and project management, where valuable knowledge surfaces but is easily lost in message streams. Documentation can preserve knowledge, but it requires ongoing maintenance and is challenging to navigate. Drawing on formative interviews that revealed organizational memory challenges in labs, we designed CHOIR, an LLM-based chatbot that supports organizational memory through four key functions: document-grounded Q&A, Q&A sharing for follow-up discussion, knowledge extraction from conversations, and AI-assisted document updates. We deployed CHOIR in four research labs for one month (n=21), where the lab members asked 107 questions and lab directors updated documents 38 times in the organizational memory. Our findings reveal a privacy-awareness tension: questions were asked privately, limiting directors' visibility into documentation gaps. Students often avoided contribution due to challenges in generalizing personal experiences into universal documentation. We contribute design implications for privacy-preserving awareness and supporting context-specific knowledge documentation.

Paper Structure

This paper contains 58 sections, 6 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. From Repository Models to Distributed Organizational Memory
  4. Leveraging Conversational Platforms for Knowledge Management
  5. AI-Powered Knowledge Management: From Chatbots to LLMs
  6. Summary and Connection to CHOIR
  7. Formative Study
  8. Method
  9. Result
  10. Challenges in Creating and Maintaining Lab Documentation
  11. Need for Effective Knowledge Retreival
  12. Importance of Interpersonal Communication
  13. Documentation as a Collective Effort with Barriers
  14. CHOIR: Chat-based Helper for Organizational Intelligence Repository
  15. Document-grounded Q&A with CHOIR
  16. ...and 43 more sections

Figures (6)

  • Figure 1: A questioner may ask CHOIR and receive a document-grounded answer, optionally share the Q&A exchange for broader discussion, and from these interactions, knowledge can be extracted and reviewed by the manager. Approved updates are written back to the organizational memory repository, which also supplies sources for Q&A.
  • Figure 2: Document-grounded Q&A with CHOIR (DM example). Bob (questioner) asks CHOIR a question via direct message, CHOIR returns a document-grounded answer, and the thread is opened to view cited references.
  • Figure 3: Two share modals in CHOIR. Left: "Share with Q&A" for posting a Q&A exchange to the designated channel. Right: "Ask in Private" for sending the same exchange via DM; in this example, the questioner adds a follow-up, the manager is pre-selected in the People picker, and anonymous sharing is enabled.
  • Figure 4: Example of knowledge extraction after a Q&A is shared. The Q&A exchange has been posted to the designated channel (#qna) through the Share with Q&A modal. A colleague (Alice) adds a follow-up comment and then mentions CHOIR to request an update, prompting CHOIR to extract documentable knowledge from the conversation.
  • Figure 5: Manager-facing document update workflow in CHOIR. Left: a member's suggestion (from Alice) arrives as a Slack DM with attribution and a link to the original discussion. Right: after starting the update process, the manager reviews the proposed change in context and decides how to incorporate it into the documentation.
  • ...and 1 more figures