Conference PaperPDF Available

Communicating and Transacting with Chatbots: Insights from Public Transport


Abstract and Figures

Chatbots as a new information, communication and transaction channel enable businesses to reach their target audience through messenger apps like Facebook, WhatsApp or WeChat. Compared to traditional chats, bots are not handled by human persons, but software is leading through conversations. Latest chatbots developments in customer services and sales are remarkable. However, in the field of public transport, little research has been published on chatbots so far. With chatbots, train customers find out timetables, buy tickets and have a personal, digital travel advisor providing real-time and context-relevant information about trips. Chatbots collect and provide different data about users and their journey in public transportation systems. They include travel, product, service and content preferences, usage patterns, demographic and location-based data. Chatbots have different advantages and high potential for both companies and mobile users. They enable new user touch points, reduce service, sales and support costs, one-to-one marketing, new data collections and deep learning. Using chatbots, smartphone users can reach a company anytime and anywhere. The questioned users of a chatbot prototype are remarkably open to new mobile services and they quickly adapt to this innovative technology.
Content may be subject to copyright.
Darius Zumstein and Sophie Hundertmark
Institute of Communication and Marketing, Lucerne University of Applied Sciences
Zentralstrasse 9, 6002 Lucerne, Switzerland
Chatbots as a new information, communication and transaction channel enable businesses to reach their target audience
through messenger apps like Facebook, WhatsApp or WeChat. Compared to traditional chats, bots are not handled by
human persons, but software is leading through conversations. Latest chatbots developments in customer services and
sales are remarkable. However, in the field of public transport, little research has been published on chatbots so far. With
chatbots, train customers find out timetables, buy tickets and have a personal, digital travel advisor providing real-time
and context-relevant information about trips. Chatbots collect and provide different data about users and their journey in
public transportation systems. They include travel, product, service and content preferences, usage patterns, demographic
and location-based data. Chatbots have different advantages and high potential for both companies and mobile users.
They enable new user touch points, reduce service, sales and support costs, one-to-one marketing, new data collections
and deep learning. Using chatbots, smartphone users can reach a company anytime and anywhere. The questioned users
of a chatbot prototype are remarkably open to new mobile services and they quickly adapt to this innovative technology.
Chatbots, chat, bots, messenger services, digital communication, digital customer services, conversational commerce
1.1 The Rise of Chatbots
The number and variety of chatbots strongly increased in the last couple of years. Today, more than 34’000
chatbots are available in the Facebook messenger (Statista 2017), and the potential global annual revenue
generated by chatbot transactions is estimated up to 32 billion US Dollars (Business Insider 2017a).
However, chatbots as a new, personal, interactive and disruptive information, communication and
transaction channel not only generate high revenues, they also reduce costs: The potential annual salary
savings created by chatbots is about 12 billion US Dollars in insurance sale, 15 billion for financial services
and sales representatives and even 23 billion for common customer service personnel in the US market
(Business Insider 2017b).
1.2 Outline and Research Method
To better understand the use, importance and the challenges of chatbots in the context of public transport, the
research project and this paper is divided into four sections: After a motivating introduction, the term chatbot
is defined and its technical functioning explained in section 2. Section 3 outlines the benefits and challenges
of chatbots. In the main section 4, the research results of chatbots in the public transport sector are presented.
The empirical study was separated in two online surveys: The first focused on 134 customers in the public
transport sector and asked about their general preferences and habits queries the timetable and to buy tickets.
Moreover, questions about the potential of a digital travel advisor including personalized information and
offers were asked. In the second survey, 84 test users of an innovative prototype were asked about their
experiences with the new chatbot. Furthermore, the potential for a travel advisor or other personalized offers
of a public train company was evaluated. Finally, section 5 of this paper gives a conclusion and an outlook.
International Conferences WWW/Internet 2017 and Applied Computing 2017
2.1 Definition
The word "chatbot" consists of the terms "chat" and "robot". Originally, the term "chatbot" was used for a
computer program, which simulates human language with the aid of a text-based dialogue system. Chatbots
contain a text input and output mask, which allows mobile users to communicate with the software behind
them, giving them the feeling of chatting with a real person (Wang & Petrina 2013).
Since the introduction of smartphones and mobile applications (apps), the term "chatbot" is increasingly
used for messenger apps rather than for pure computer programs (Atwell & Bayan 2015).
2.2 Operating Mode of Chatbots
Generally, chatbots have quite similar technologies and architectures. Figure 2 shows the technical process of
a chatbot, when a mobile user makes a request until the appropriate answer is sent by the chatbot.
Figure 1. Operating mode of chatbots (Source: Following Weidnauer 2016)
The process starts with a user's request (see step 1 in Figure 1) using a messenger app like Facebook,
Slack, WhatsApp or Skype, or an app using text or speech input (e.g. Amazon Echo in step 2). The user
request is recorded by a so-called Natural Language Parser (NLP; 3) and is translated into the programming
language of the conversation engine. Following, the conservation engine analyses the question and redirects
it to the backend (4). The backend is linked to one or several databases (DB), which give a result to the
corresponding query. Once the appropriate result has been found in the backend (5 and 6), the conversion
engine forwards it to the response picker (7). Now the answer, which is still in the programming language of
the chatbot, is translated into the natural language of the user and sent to him to user interface (see 8 and
Weidnauer 2016).
Chatbots and language parsers use semantic patterns and keywords to analyze the requests of users and to
edit them as accurately as possible. By matching databases stored in the backend, chatbots recognize patterns
or regularities and combine them. This process is also called machine learning. In addition, many chatbots
use the technique of deep learning, a subcategory of machine learning.
The chatbot starts with the analysis of the main core themes and then goes into the depth of the topic. If
the user starts the conversation with a question, the chatbot first tries to analyze the main topic and then uses
the funnel principle to narrow down the topic more closely (Dempt 2016). Software tries to understand the
text of the user by regular and data-driven semantic procedures. Rule-based methods attempt to automatically
recognize data expressions. Data-oriented methods work similarly to content analysis of qualitative social
research. Deductive categories are built in advance and then the texts of the users are coded using these
categories to assign them quickly to the related topics (Trendone 2016).
ISBN: 978-989-8533-69-2 © 2017
As soon as the chatbot has solved the user's request, it has to provide the right answers as quickly as
possible. For this purpose, various database entries are stored in the backend (see figure 1 on the right): These
can be Customer Relationship Management (CRM) systems, Enterprise Resource Planning (ERP) systems,
Content Management Systems (CMS), Product databases or Information Systems (PIM) or another internal
or/and external database. The chatbot matches the given question with the databases in the backend and
provides the user with the requested information or suggestions (Dempt 2016).
2.3 Application Fields of Chatbots
The fields of applications of chatbots are manifold: Very popular are calendar assistants and chatbots for
reserving or purchasing event tickets (for 68% and 64% of the responses in figure 2). Searching and buying
products online using chatbots are very popular too (e.g. H&M, chatShopper & eBay). Moreover, 58% of the
asked users are using chatbots for booking hotels, trips and flights (e.g. KLM & Austrian Airline chatbot).
This paper focuses on chatbots used for searching and booking train trips. Other fields are chatbots for
news (for instance the CNN chatbot), weather (e.g. Hi Puncho), traffic (e.g. Traffic News & Traffic Jam) and
financial chatbots (e.g. Trading Bot). Finally, many chatbots are used for customer and delivery services.
Figure 2. Fields of applications of chatbots (Source: Statista 2017b)
3.1 Strengths and Benefits of Chatbots
Chatbots have two different type of consequences for companies: On the one hand, chatbots change the way
of informing, communicating and transacting between the company and its customers or other stakeholders.
On the other hand, they influence the organization, communication and collaboration within the company.
Thanks to messenger apps and chatbots, most businesses have new ways to interact with their customers
through one-to-one communication. Users usually use the messenger apps for private purposes among friends
and colleagues. Companies have now the chance to enter this private communication channel for businesses.
Using chatbots, consumers and businesses can communicate 24 hours day, 7 days the week, independent of
working or opening hours. Companies can save on personnel costs in customer services and do not take the
risk that they cannot be reached outside their business hours, and they do not miss customers’ requests.
Looking at the collection of user and usage data, the use of chatbots leads to new potentials for providers.
Companies get to know their customers in a new way. In many cases, users link their social media profiles
with their messenger profiles, so that companies can get direct insights into user interests, responses and
profiles. If this is not the case, the chatbot can send the necessary information or questions during the dialog
with users. In addition, the chatbot stores individual user preferences based on the users purchase history,
inquiries and other activities. These new data collections give companies the opportunity to address their
customers in a targeted manner, and the customized offers can be targeted directly and personally to users.
0% 20% 40% 60% 80%
Calender assistants
Research during online shopping
Booking of hotel, trips, flight & train tickets
Daily news (e.g. weather, news, traffic, stock)
Customer service
Delivery service
Share of
Focus of
this paper
International Conferences WWW/Internet 2017 and Applied Computing 2017
Apart from the advantages of communication and transaction, chatbots also offer new potentials within
the company. For example, chatbots are used for supporting and training purposes for employees. Another
example is the digital onboarding of new staff members. Instead of a long onboarding of colleagues, chatbots
take over the introduction of new team members, internal communication, standard processes and tasks.
Chatbots allow customers to get in contact with companies whenever they want so, without having to pay
attention to time zones and to opening times and waiting loops of call/service centers. Chatbots are promising
for international and digital companies like online retailers or web shops. Customers sometimes buy products
in different countries and they do not want to be dependent to local time zone or foreign languages. A further
advantage is personal one-to-one communication. So far, users often must search and browse a website for a
long time to find the right information like product, price, service or contact information. In the case of
complaints or other customer service inquiries, chatbots are helpful, straight-forward and efficient.
In a best-case scenario, the chatbot knows its users like a good friend and offers them appropriate offers,
solutions and services at the right time. Depending on the permissions the user gives to the chatbot, he will be
informed automatically and proactively on specific inquiries and demands. In public transport, the chatbot as
a travel companion is a promising service compared to other ones. In future, customers can automatically be
informed by push notifications about delays and other relevant information directly on users’ mobile phone.
3.2 Weaknesses and Threats of Chatbots
In contrast to the many benefits of chatbots, researchers, developers and providers of chatbots should also be
aware of the disadvantages and risks of the new applications. Today, customers are very familiar to receive
or request information from companies using phones, newsletters, e-mails, apps or websites. Messengers are
mainly used for private communication among friends and colleagues.
However, companies must be aware that customers are used to other communication channels and that it
will therefore take some time to adapt before getting used to the new communication and buying methods. In
a transition phase, classical (offline and online) channels still must be provided and the customers should be
motivated and incentivized to use the new tools. It is important that companies can be found in messenger
apps like Facebook or, in the case of standalone apps, in the Apple and Google app store.
Another important topic for both providers and users is data protection. If companies offer a stand-alone
chatbot app, they are responsible for protecting and handling customer data adequately. This is just like with
traditional apps and websites. However, if companies offer their chatbot on a third-party platform, data are
also sent to operators and platforms like Facebook, WhatsApp, Slack or WeChat. Chatbot developers and
operators should ensure both data privacy and data protection. When it comes to payment processes, where
bank or credit card data are insert, data protection is crucial. Communicating with customers, companies try
to collect as many user data as possible, to store them and to use them for further transactions or marketing
offers. Users need to be aware that providers of chatbots and messenger platforms will collect personal data.
Furthermore, some consumers may fear that they will miss other offers because of personalized offers are
selected by a bot only. To reduce this risk, users should ask chatbots for other offers outside of their
preferences. Table 1 summarizes some important strengths/opportunities and risks of chatbots.
Table 1. Advantages and risks of chatbots for providers and users
Strengths & Opportunities
Weaknesses & Risks
For providers/
- 24/7 customer service (anytime/anywhere)
- New & direct customer contact points
- New methods & types of data collection
- High amount of personal user/usage data
- Personalization & automation of communication
- Reduction of service & support costs
- Malfunctioning chatbots & unanswered questions
- Inve stments i n IT infrastructure & chatbot tools
- Extension of IT & analytics architectures
- Lack of awareness & acceptance by users
- Information security & data protection
- Image & reputation risks
For users/
- 24/7 customer services & support
- One-to-one communication on personal device
- High convenience & ease of use
- Time- & cost-savings
- Reduction on relevant information & services
- Relevant offers based on user preferences
- Privacy
- Data protection of personal & sensitive data
- Lack of experience & understanding
- Biased personalized information
- Artifi cial/non-human conversation
- Social isolation & ethical concerns
ISBN: 978-989-8533-69-2 © 2017
4.1 The New Chatbot of the Swiss Railway Company BLS
Since the beginning of April 2017, the chatbot of the Swiss Railway Company BLS is publicly available in
the Apple app store (an Android version in Google play is following soon). Previously, it was treated as a
prototype and was available for internal developing and testing only.
The first chatbot version of BLS allows its users to buy individual tickets, day tickets in Switzerland and
additional tickets in the BLS tariff area. Furthermore, it offers the possibility to simple query the timetable of
all trains. The user can request both core functions by chatting with the bot within the native app.
Unfortunately, the conversation with the chatbot is quite static now. The customers cannot enter free
texts, but choose from various response options only. Furthermore, the chatbot is not yet functioning as a
permanent travel companion. In future, the chatbot will be able to process free, imitated conversations by the
user and provide regular information on travel histories or other travel-related offers like a travel companion.
4.2 Methods of Ticket Purchases and Experience with Chatbots
In the first survey, following question asked was asked to 134 participants: "Which method do you currently
use most frequently to buy your tickets?" (multiple answered where possible). The responses showed that
almost every second customer still is using the ticket machine to purchase tickets and 43% use an app to buy
tickets (compare Figure 3a). However, every fifth passenger still uses the ticket counter at the train station,
where they have personal contact with a sales and service staff member.
To purchase train tickets, the website is used by 12% of customers only. Obviously, passengers are used
to use apps before, during or after their journey. In the last years, websites as point of sales lost in importance
in public transports. Finally, travel centers are also very seldom visited to purchase tickets (2%).
The results show that ticket machines, which are available at every train stations, and apps are the both
most widely used methods to buy tickets. Digital channels as the (mobile) website, which are mostly used at
home or at the office, are less and less used for buying tickets.
Figure 3. a) Methods of ticket purchase in public transport and b) experience with chatbots (n=134)
Secondly, in the participants of survey were asked, if they are familiar with chatbots or if these
applications are (completely) new to them. Figure 3b shows that most participants already know chatbots:
28% of the high experienced respondents are using chatbots in other application fields. 40% have at least
some experience and used chatbots sometimes. 24% have few experience with chatbots and did not used
them often. 8% of the users had never heard about the term chatbot”. These responses show that chatbots are
not completely new to customers, but this type of applications is already popular.
The answers reflect and confirm the results of the literature: Chatbots are not new to most of people, but
just a minority is using them regularly and intensively. Furthermore, literature research shows that chatbots
generally get great acceptance from most users. Regarding the survey, it is certainly helpful for many mobile
users that they are familiar with chatbots.
However, many mobile users and customers still must learn this new information and communication
system. Public transport customer mobile user in general must adapt their habits, their searching and
buying behavior to chatbots, which takes some time, learning, practice and communication efforts.
0% 10% 20% 30% 40% 50%
Ticket machine
Mobile app
Ticket counter at station
Travel center
International Conferences WWW/Internet 2017 and Applied Computing 2017
4.3 Digital Services and Real-Time Information in Public Transport
Considering the registration and onboarding process of the BLS chatbot, the test users were happy: 80% were
(very) satisfied (in figure 4a). The timetable and ticket access provided by the bot also have seen as positive
and there was no problem with the payment process. However, the payment process and especially the ticket
offering should and will be improved in a future version of the chatbot.
In future, a chatbot in public transport will serve as a travel companion, as a personal digital consultant.
To find out whether the users would like travel support, the participants were asked whether they want real-
time information during the trip, such as delays or transfers. Figure 4b shows that customers answered very
united to this question: 84% of mobile users always want real-time information and updates about their trips.
12% of the respondents would like to receive information only if they explicitly ask for it, and insignificant
4% have indicated that they do not want to receive real-time information at all. One reason for the high level
of consensus is the fact that other companies are already providing order or delivery status information, e.g.
e-mail or SMS notifications about a customers’ mail, parcel or product. After completing a payment or
buying process, customers are asked whether they wish to be informed by e-mail, SMS or by a push
notification in mobile apps, for instance in the case of train delays, next connections or changes of the rail
track. The chatbot will provide real-time and highly relevant information to its user in the form of chat
messages. This means that the customer does not have to enter his e-mail address or phone number, but gets
the information directly from the chatbot as a push notification referring directly to text in the messenger app.
The comparison with other providers and the detailed analysis of the prototype showed that the chatbot
should not provide timetables and tickets only, but also serves as a personal travel companion, as a robot
adviser providing context- and location-based information during a customer journey in public transport.
Figure 4. a) User satisfaction with the BLS chatbot and b) demand for real-time travel information (n=84)
4.4 Future Usage of Chatbots in Transportation
Users will continue to use a (chatbot) prototype only, if she/he is satisfied after an initial test phase and if the
new solution provides high customer (added) value. Therefore, the participants were asked at the end of the
online survey, whether they will continue to use the chatbot for buying tickets or for timetable information.
The answers in Figure 5a show that the chatbot prototype has convinced many, but not all users: 40% of
respondents would continue to use it. In contrast, 48% are not quite sure and probably prefer other methods.
12% will switch back to the previous method of searching trains and buying tickets (using common apps or
the ticket machines at the station). Considerin g the answers of the previous questions, this result is plausible.
Figure 5. a) Future usage of chatbots in transportation and b) chatbots in apps or in messengers (n=84)
0% 20% 40% 60% 80%
Ticket access
Payment methods
Ticket offers
Very satisfied
ISBN: 978-989-8533-69-2 © 2017
On the one hand, the users miss the entry of different ticket preferences and on the other hand, the chatbot do
not offer all kind of tickets available. Since chatbots like the BLS prototype still have functional or technical
deficits and still are in their infancy, many users will not (yet) use the chatbot regularly. At the same time, the
chatbot hast to be improved and extended so that customers can use it for all inquiries and receive value.
A chatbot can be provided as a standalone app or as an integration into a messenger app like the Facebook
messenger, WhatsApp or Slack. Th erefore, respondents were asked if they prefer to use the chatbot as a
separate app or as a chatbot in a common messenger. Figure 5b shows that users have a clear opinion to the
this question: 71% prefer the chatbot as a stand-alone app. 13% of the participants wish to integrate it into
WhatsApp and 8% prefer an integration in the Facebook messenger only, probably for privacy reasons.
This results differ from literature research. This states that the trend goes to messenger apps like
Facebook messenger or WhatsApp, and away from apps. In most European (national or local) public and
private transport systems, customers are used to use a mobile app to looking for time tables and to buying the
next flight, train, metro or bus ticket, often displayed in the wallet of the iPhone or Android mobile phone.
Today, thousands of organizations like the Swiss Railways companies BLS and SBB are developing and
providing chatbots for mobile smartphones. In future, chatbots can complement or even replace traditional
information, communication and sales channels like newsletters, websites, sales desks or hotlines.
With chatbots, both companies and users can initially, directly and personally contact at anytime and
anywhere, what is not possible in this manner with traditional channels like websites, newsletter or hotlines.
In addition, chatbots can communicate context- and location-based to prospects and existing customers
and allow personalization as well as one-to-one marketing (see table 2).
Table 2. Chatbots versus traditionnel communication channels
Social Media
Contacting by companies
Contacting by users or customers
Customer service outside the service hours
Communication to new customers (acquisition)
Communication to customers (retention)
Communication among users
Machine/deep learning (Artificial Intelligence)
Personalization / 1-to-1 marketing
Context-/location-based services
Legend: ++ st ren gth + possible difficult weakness (not possible)
However, it must be ensured in digital marketing and digital business that chatbots can understand and
edit most of user requests without human help. If this is not possible, a customer consultant should be
available in the background and service the customer. Native apps and (mobile) websites will continue to
persist despite the rise of chatbots and still be used by mobile users. Some of the app features and
functionalities probably will be replaced by chatbots and other intelligent assistants like Siri or Amazon
However, a complete replacement of apps, websites and web shops cannot be observed so far. Most users
are still used and bound to conventional communication and transactional channels like mobile apps and
websites. This setting will probably continue for a couple of years. Since current chatbots still are not able to
adequately cover all features, chatbot technologies and their applications should be developed and improved
by researchers, by the software industry and by digital businesses. The same applies to social media channels:
They offer many functions that chatbots currently do not cover yet and it is questionable whether they will
provide them in near future. Thus, corporate social media and chatbots will co-exist and used by digital users.
Nevertheless, chatbots and their markets are rising, new fields of applications and usage are growing very
fast. Companies and researchers must invest, investigate, test and develop chatbots further. Chatbots are one
of the next generation intelligent information and communication systems which make our lives easier and
more convenient.
International Conferences WWW/Internet 2017 and Applied Computing 2017
Atwell, E., Bayan, A.-S. (2015). ALICE Chatbot: Trials & Outputs. In: , vol. 19.
Business Insider (2017a). Here's how chatbot metrics differ from traditional apps, BI Intelligence, Available:, accessed 24th of June
Business Insider (2017b). Facebook has opened up analytics and developer tools for Messenger bots, BI Intelligence
accessed 24th of June 2017.
Christensson, P. (2009), Avatar, Available:, accessed 24th of June 2017
Coniam, D. (2008). Evaluating the language resources of chatbots for their potential in English as a second language.
In: ReCall (20), pp. 98-116.
De Pietro, G., Gallo, L., Howlett, R., & Jain, L. (2016). Smart Innovation, Systems and Technologies. Switzerland:
Springer International Publishing Switzerland.
Dempt, F. (2016). Futuregram. Available:, accessed 24th of June
Gill, A., & Oberlander, J. (2002). Taking Care of the Linguistic Features of Extraversion, Availble:
cogscilfinaLind_files/gilLOberlander.pdf, accessed 24th of June 2017
Hill, J., Ford, W., & Farreras, I. (2015). Real conversations with artificial intelligence: A comparison between
humanhuman online conversations and humanchatbot conversations. In: Computers in Human Behavior (49), pp.
Jia, J. (2009). CSIEC: A computer assisted English learning chatbot based on textual knowledge and reasoning.
In: Knowledge-Based Systems, pp. 249-255.
Laurel, B. (1997). Interface Agents: Metaphors with Character. Software Agents, pp. 67-77.
McNeal, M., & Newyear, D. (2013). Chatbot Creation Options. Library Technology Reports (49).
Mena, J. (2012). Machine-to-Machine Marketing (M3) via Anonymous Advertising Apps Anywhere Anytime (A5),
Auerbach Publications, New York.
Reeves, B., & Nass, C. (1996). The media equation: how people treat computers, television and new media like real
people and places. Cambridge University Press, New York.
Statista (2017a). Statista, VentureBeat, 11/2016. The number of chatbots for the Facebook Messanger. Available under:
messenger/, accessed 24th of June 2017.
Statista (2017b). Statista, Bitkom, 11/2016. For which fields would you use chatbots? Available under:
deutschland, accessed 24th of June 2017.
Trendone (2016). Futuregram. Available:, accessed 24th of June 2017.
Van Lun, E. (2016). Chatbots. Available:, accessed 24th of June 2017.
Wang, Y., & Petrina, S. (2013). Using Learning Analytics to Understand the Design of an Intelligent Language Tutor.
In: International Journal of Advanced Computer Science & Applications (11), pp. 124-131.
Weizenbaum, J. (1966). ELIZA - A Computer Program For the Study of Natural Language Communication Between
Man And Machine. In: Communications of the ACM (9).
Woods, J., & Abdul-Kader, S. (2015). Survey on Chatbot Design Techniques in Speech Conversation Systems.
In: International Journal of Advanced Computer Science and Applications (7).
Zumstein, D., & Hundertmark, S. (2017). Chatbots are Coming Opportunities and Risks for Online Marketing
(in German), In: Online Marketing Conference, 17th of August 2017, Bern.
ISBN: 978-989-8533-69-2 © 2017
... In our corpus, a total of 12% of the papers (N=10) tackled topics related to the acceptability and acceptance of chatbots. Three papers focus on the assessment of the user acceptance of specific newly developed chatbots, in the teaching (de Medeiros et al., 2019), company (Fiore et al., 2019) and transportation (Zumstein & Hundertmark, 2017) contexts. The other papers, instead, explore wider issues, trying to identify either the reasons why people are open to accept this technology (5 papers), or those motivations that lie behind its use (2 papers). ...
... A considerable number of papers (16) in our corpus aim to understand and assess the user's 'satisfaction' when interacting with a chatbot (e.g., Liu & Dong, 2019;Mendez et al., 2019;Procter et al, 2018;Galko et al., 2018;Zumstein & Hundertmark, 2017;Jin et al., 2019). HCI regards satisfaction with a technological artifact as a major design goal (e.g., ISO 1998). ...
Over the last ten years there has been a growing interest around text-based chatbots, software applications interacting with humans using natural written language. However, despite the enthusiastic market predictions, ‘conversing’ with this kind of agents seems to raise issues that go beyond their current technological limitations, directly involving the human side of interaction. By adopting a Human-Computer Interaction (HCI) lens, in this article we present a systematic literature review of 83 papers that focus on how users interact with text-based chatbots. We map the relevant themes that are recurrent in the last ten years of research, describing how people experience the chatbot in terms of satisfaction, engagement, and trust, whether and why they accept and use this technology, how they are emotionally involved, what kinds of downsides can be observed in human-chatbot conversations, and how the chatbot is perceived in terms of its humanness. On the basis of these findings, we highlight open issues in current research and propose a number of research opportunities that could be tackled in future years.
... Another study on chatbots compared the capabilities of general offline customer service agents in the marketing field and how to represent brands and improve customer/brand relationships [3]. Besides, the findings obtained by [13] from an organizational perspective show that chatbots can significantly affect and change the organisation's future, communication, and collaboration within companies. While research conducted by [11], using factors to accept customer satisfaction aspects such as usability, brand image, personality, and ease of use for chatbot adoption in Indonesia [11]. ...
Full-text available
the goal of this article is to explore how learning analytics can be used to predict and advise the design of an intelligent language tutor, chatbot Lucy. With its focus on using student-produced data to understand the design of Lucy to assist English language learning, this research can be a valuable component for language-learning designers to improve second language acquisition. In this article, we present students’ learning journey and data trails, the chatting log architecture and resultant applications to the design of language learning systems.
Full-text available
This paper investigates the linguistic worth of current ‘chatbot’ programs – software programs which attempt to hold a conversation, or interact, in English – as a precursor to their potential as an ESL (English as a second language) learning resource. After some initial background to the development of chatbots, and a discussion of the Loebner Prize Contest for the most ‘human’ chatbot (the ‘Turing Test’), the paper describes an in-depth study evaluating the linguistic accuracy of a number of chatbots available online. Since the ultimate purpose of the current study concerns chatbots' potential with ESL learners, the analysis of language embraces not only an examination of features of language from a native-speaker's perspective (the focus of the Turing Test), but also aspects of language from a second-language-user's perspective. Analyses indicate that while the winner of the 2005 Loebner Prize is the most able chatbot linguistically, it may not necessarily be the chatbot most suited to ESL learners. The paper concludes that while substantial progress has been made in terms of chatbots' language-handling, a robust ESL ‘conversation practice machine’ (Atwell, 1999) is still some way off being a reality.
Full-text available
Can human beings relate to computer or television programs in the same way they relate to other human beings? Based on numerous psychological studies, this book concludes that people not only can but do treat computers, televisions, and new media as real people and places. Studies demonstrate that people are "polite" to computers; that they treat computers with female voices differently than "male" ones; that large faces on a screen can invade our personal space; and that on-screen and real-life motion can provoke the same physical responses. Using everyday language to engage readers interested in psychology, communication, and computer technology, Reeves and Nass detail how this knowledge can help in designing a wide range of media.
CSIEC (Computer Simulation in Educational Communication) system with newly developed multiple functions for English instruction still focuses on supplying a virtual chatting partner (chatbot), which can chat in English with the English learners anytime anywhere. It generates communicative response according to the user input, the dialogue context, the user’s and its own personality knowledge, common sense knowledge, and inference knowledge. All these kinds of knowledge are expressed in the form of NLML, an annotation language for natural language text. These NLMLs can either be automatically obtained through parsing the text, or be easily authored with the help of GUI editors designed by us. So the CSIEC system suggests a naïve approach of logical reasoning and inference directly through syntactical and semantic analysis of textual knowledge. This approach has advantages over the old ELIZA-like keywords matching mechanism. The chatting log summarization of free Internet usage within six months demonstrates this advantage. In this paper, we present the system architecture and underlying technologies, and the educational application results.
We study how Extraversion or Introversion influences people's language production. A corpus of e-mail texts was gathered from individuals categorised via Eysenck's EPQ-R personality test. One experiment analysed the corpus using existing content analysis tools, and found relatively weak effects of Extraversion. A second experiment used more sensitive bigram-based techniques from statistical natural language processing to replicate earlier findings, and uncover novel patterns of behaviour.
  • E Atwell
  • A.-S Bayan
Atwell, E., Bayan, A.-S. (2015). ALICE Chatbot: Trials & Outputs. In: , vol. 19.