Designing Adaptive Developer-Chatbot Interactions: Context Integration, Experimental Studies, and Levels of Automation

TL;DR

This paper explores how integrating contextual information into AI chatbots can improve support for software developers, through experimental studies on interaction design and automation levels.

Contribution

It introduces a novel approach to designing adaptive, context-aware chatbots for software developers, with insights from experimental user studies.

Findings

Enhanced understanding of developer expectations from chatbots

Insights into optimal levels of automation for developer support

Design principles for context integration in chatbots

Abstract

The growing demand for software developers and the increasing development complexity have emphasized the need for support in software engineering projects. This is especially relevant in light of advancements in artificial intelligence, such as conversational systems. A significant contributor to the complexity of software development is the multitude of tools and methods used, creating various contexts in which software developers must operate. Moreover, there has been limited investigation into the interaction between context-based chatbots and software developers through experimental user studies. Assisting software developers in their work becomes essential. In particular, understanding the context surrounding software development and integrating this context into chatbots can lead to novel insight into what software developers expect concerning these human-chatbot interactions and their levels of automation. In my research, I study the design of context-based adaptive interactions between software developers and chatbots to foster solutions and knowledge to support software developers at work.