Page 57 - DP-2-3

Design+ Design chatbot using activity theory

traffic in Malaysia, traffic road signs guarantee a smooth By optimizing operations and streamlining actions,
and safe driving experience for all drivers. chatbots can enhance the UX.
Despite advancements in GPS navigation, studies 2.2.3. Contradictions and breakdowns
have shown that drivers’ over-reliance on GPS can lead
to reduced awareness of their surroundings, increasing Activity theory recognizes that contradictions (conflicts
the risk of traffic accidents due to unexpected situations within the system) can lead to breakdowns. For chatbots,
or navigation system failures. The over-reliance on GPS these breakdowns might occur when the chatbot fails
to understand user inputs or when the interface is not
has induced the degeneration of memory and poor culturally appropriate. Identifying and resolving these
reconstruction of the environment among the GPS users. contradictions–such as integrating multilingual support
The distraction brought by the GPS could also cause traffic or localizing chatbot interfaces–can significantly improve
accidents to happen. In addition, most signs utilize pictures usability. 8
rather than words, so it may lead to misunderstanding.
One avenue to explore in assisting the users to interpret 2.2.4. Cultural-historical context
traffic road signs was to develop the Expert Systems Usability challenges can also stem from the chatbot’s lack
prior to upgrading the system to generative artificial of alignment with users’ cultural norms and expectations.
intelligence. An Expert System is a computer program Activity theory emphasizes designing tools that are
that simulates the judgment and behavior of a human culturally relevant and historically informed. For instance,
with expert knowledge and expertise in a particular field. a cultural heritage chatbot could incorporate stories and
It is very popular across different domains, such as in terminologies familiar to its target audience to create a
diagnostic medicine (e.g., arthritis, disorders, agriculture, more meaningful UX. 36
and information libraries). 14
2.3. Design of road signal chatbot using activity
This case study presents the implementation of an
Expert System enhanced with NLP capabilities, enabling theory
users to interact with the system conversationally through There are two methods that we can use to design chatbot.
a chatbot. The chatbot delivers real-time traffic and route First, it is possible to select suitable concepts from AT that
information via an interactive interface, aiming to educate are deemed relevant for usability and UX design for the
users about road signs while ensuring a user-friendly and road sign chatbot. The crucial consideration is that the
engaging experience. selected concepts should be able to guide both the data
gathering and analysis process and transfer results into a
2.2. Key concepts of activity theory applied to design representation with structure and continuity. The
chatbot design second is to use the expanded triangle model of activity
systems of Engstrom to capture concepts from activity
2.2.1. Subject-tool-object framework theory that are relevant to the analysis of work practices
Activity theory views tools (chatbots) as mediators between and chatbot design, whilst giving a structured approach to
the subject (user) and the object (goal). For chatbots, the analysis.
usability issues often arise when the tool fails to mediate Activity theory considers the road signal chatbot as a
effectively due to poor interface design, lack of intuitive special kind of tool mediating human interaction with the
interactions, or insufficient context-awareness. Designing world. Users use the road signal chatbot not because they
chatbots that dynamically adapt to user goals and contexts want to interact with it, but because they want to reach
can bridge this gap. For instance, a chatbot in education their goals beyond the situation of the “dialogue” with the
could offer context-specific prompts based on the learner’s chatbot. Bødker and Grønbæk called this, “users acting
29
progress. 8 through the interface.” The road signal chatbot must be
36
usable and enjoyable for the users. This is concerned with
2.2.2. Hierarchy of activity
usability and UX.
Activities are structured in three levels: Activities (high-
level goals), actions (intermediate steps), and operations 2.3.1. Clarify the purpose of the activity
(automatic processes). Chatbot usability can be improved The first step in the design is to clarify the purpose of the
by mapping user interactions to this hierarchy. For example: activity. It is important to clarify the motives and goals
(i) Activity: A user wants to learn about road signs. of the activity system. The reasons are to understand
(ii) Actions: Searching for particular road sign to learn. the context within which activities occurred to reach a
(iii) Operations: Auto-filling personal details. thorough understanding of the motivation for the activity

Volume 2 Issue 3 (2025) 7 doi: 10.36922/DP025060009