[Paper Review] Evaluating Quality of Chatbots and Intelligent Conversational Agents

Nicole Radziwill, Morgan C. Benton|arXiv (Cornell University)|Apr 15, 2017

AI in Service Interactions23 references184 citations

TL;DR

The paper surveys quality issues and attributes for chatbots, reviews quality assessment approaches, and proposes an Analytic Hierarchy Process–based method for evaluating chatbot quality.

ABSTRACT

Chatbots are one class of intelligent, conversational software agents activated by natural language input (which can be in the form of text, voice, or both). They provide conversational output in response, and if commanded, can sometimes also execute tasks. Although chatbot technologies have existed since the 1960s and have influenced user interface development in games since the early 1980s, chatbots are now easier to train and implement. This is due to plentiful open source code, widely available development platforms, and implementation options via Software as a Service (SaaS). In addition to enhancing customer experiences and supporting learning, chatbots can also be used to engineer social harm - that is, to spread rumors and misinformation, or attack people for posting their thoughts and opinions online. This paper presents a literature review of quality issues and attributes as they relate to the contemporary issue of chatbot development and implementation. Finally, quality assessment approaches are reviewed, and a quality assessment method based on these attributes and the Analytic Hierarchy Process (AHP) is proposed and examined.

Motivation & Objective

• Identify and categorize quality issues and attributes relevant to contemporary chatbots and conversational agents.
• Review existing quality assessment approaches and frameworks for chatbots.
• Propose and preliminarily examine a quality assessment method based on identified attributes and Analytic Hierarchy Process (AHP).

Proposed method

• Conduct a literature review of quality issues and attributes in chatbot research.
• Synthesize quality attributes and assessment practices from prior work.
• Develop a proposed quality assessment method grounded in the identified attributes and the Analytic Hierarchy Process (AHP).
• Provide examination of the proposed method within the context of chatbot quality evaluation.

Experimental results

Research questions

• RQ1 What quality issues and attributes are relevant to modern chatbots and intelligent conversational agents?
• RQ2 What quality assessment approaches exist for evaluating chatbots, and how effective are they?
• RQ3 How can an AHP-based method be used to assess chatbot quality based on identified attributes?
• RQ4 What does preliminary examination of the proposed method suggest about its utility for quality evaluation?

Key findings

• The literature review identifies specific quality issues and attributes related to chatbot development and deployment.
• Various quality assessment approaches for chatbots are reviewed and summarized.
• A quality assessment method based on identified attributes and the Analytic Hierarchy Process (AHP) is proposed and examined.
• The study links quality attributes to practical evaluation strategies for conversational agents.