Page 52 - DP-2-3

Design+ Design chatbot using activity theory

personalized, on-demand instruction, making learning These limitations can result in a lack of trust, low
more engaging and accessible. However, the success of engagement, and reduced effectiveness in achieving the
such systems hinges on their usability. User experience intended goals.
(UX) issues can also impede learning outcomes and user Both usability and UX influence how easily and
satisfaction. satisfactorily users can interact with the system because

1
Manawadu and Wijenayake developed a voice- they are crucial for the design of effective chatbots to
activated chatbot that utilizes convolutional neural ensure that chatbots meet user needs, foster engagement,
networks for real-time traffic sign recognition. While the and achieve their intended purpose. Usability is a critical
system offers auditory feedback, users have reported issues factor in the design of effective chatbots because it ensures
with environmental factors such as poor lighting and that users can achieve their goals efficiently, effectively, and
adverse weather conditions, which can affect the accuracy satisfactorily. Good usability enables users to accomplish
of sign recognition. In addition, the chatbot’s natural their tasks quickly, with minimal stress and errors, and
language processing (NLP) capabilities sometimes struggle ultimately feel satisfied when interacting with a product’s
with complex or ambiguous sign descriptions, leading to user interface. Besides usability, UX is also important for
misinterpretations. chatbots because it helps to create a positive interaction
between a user and a chatbot. It directly influences user
The European Union’s Erasmus + funded project “Smart satisfaction, engagement, and the overall success of
Learning for Road Safety” employs a chatbot integrated interactions. A chatbot with a well-designed UX not only
with an e-learning platform to educate users about road meets user needs but also fosters trust and long-term usage.
signs and traffic rules. Despite its interactive approach, A well-designed chatbot provides a seamless interaction
users have encountered difficulties with the platform’s experience, minimizes user frustration, and enhances task
accessibility features. completion rates. By adopting proven usability and UX
Many chatbots fail to meet user expectations due design principles, chatbots can better meet user needs and
to shortcomings in usability and UX. There are many expectations.
common usability problems in chatbot road signs in use. Theoretical frameworks provide a structured foundation
These include: for designing usability and UX. They help designers
(i) Inaccurate or irrelevant responses. Users often understand human behavior, predict user interactions, and
report that chatbots provide generic or irrelevant create systems that meet user needs efficiently. In his paper,
information, leading to frustration and diminished Norman states that applying theories to design ensures a
2
trust in the system. This issue is particularly scientific, repeatable approach to design, improving the
pronounced when chatbots fail to understand likelihood of creating effective, user-friendly chatbots.
context or user intent, resulting in responses that do
not address the user’s specific query. 1 There are several theories and models for designing
(ii) Accessibility barriers. Users with disabilities, usability and UX for chatbots. These include Norman’s
2
3
particularly those with visual impairments, often Design Principles, Nielsen’s Usability Heuristics, User-
4
face challenges when interacting with chatbots. Centered Design, The Eight Golden Rules of Interface
6
4
5
Issues such as focus jumping, where the screen Design, Emotional Design, Persuasion Technology, and
7
reader’s focus shifts unexpectedly, and inadequate Conversational UX Design Framework. Despite the fact
announcements of new messages can impede the that each of these theories provides useful principles for
user’s ability to engage effectively with the chatbot. 2 chatbots design, they lack the ability to model complex
8
(iii) Lack of natural conversational flow. Many systems because they often focus on specific aspects of
educational chatbots struggle to emulate natural usability or interface design.
human conversation, leading to interactions that feel Activity theory, with its holistic, systemic approach,
mechanical or disjointed. This lack of conversational is better suited for designing chatbots that operate in
fluidity can disengage users and reduce the dynamic, real-world contexts. It ensures that chatbots
effectiveness of the learning experience. not only provide usable interfaces but also align with the
(iv) Repetitive interactions. Some chatbots exhibit social, cultural, and organizational factors that influence
repetitive behavior, asking the same questions or user interactions. It is our belief that activity theory offers
providing identical responses without progressing a structured approach to understanding and designing
the conversation. Such loops can waste users’ time usability and UXs for effective chatbot. This study shows
and hinder the learning process, especially when the how activity theory principles can be effectively used to
chatbot does not adapt to previous inputs. design usability and UX in a chatbot for road sign users.

Volume 2 Issue 3 (2025) 2 doi: 10.36922/DP025060009