Summary
Chatbots have caught the headlines recently with businesses starting to adopt them to stimulate conversation with customers. But what are chatbots? How do they work? What can they do for museums and their audiences?
Authors
Stefania Boiano, InvisibleStudio LTD, UK
Ann Borda, University of Melbourne, Australia,
Pietro Cuomo, Art in the City, Italy
Giuliano Gaia, InvisibleStudio LTD, UK
Stefania Rossi, Museo Poldi Pezzoli, Italy
Keywords
museums, chatbots, artificial intelligence, gamification, teenagers
Introduction
Chatbots, also known as talkbots or chatterbots or bots, are growing exponentially in their use by marketers and online businesses in enhancing customer experiences, often as messaging applications that can personalise the interaction (e.g. recommender systems) (1). Simply put, chatbots are computer programs that mimic conversation using auditory or textual methods. More specifically the functionality of chatbots use natural language processing that has a history rooted in artificial intelligence (AI).
ELIZA:
One of the earliest such natural language applications was a chatbot called ELIZA developed from 1964 to 1966 at the MIT Computer Science and Artificial Intelligence Laboratory by computer scientist Joseph Weizenbaum. ELIZA originally was created to use simple pattern matching and a template-based response (prewritten scripts) to emulate the conversational style of a psychotherapist. ELIZA generated a wave of global community interest in building a conversational bot that could pass the Turing Test.
Turing Test:
Published in 1950, Alan Turing’s Computing Machinery and Intelligence addresses the overarching question: Can machines think? (2)
The Turing test in its most simple form is carried out as a sort of imitation game. The test has a human interrogator speaking to a number of computers and humans through an interface. If the interrogator cannot distinguish between the computers and the humans then the Turing Test has been passed. This quest found an audience through the Loebner Prize (begun in 1991), which has taken the form of an annual competition designed to implement the Turing Test.
A.L.I.C.E.:
Building on the pattern-matching techniques used in ELIZA and advancing natural conversational language capabilities, American scientist Richard Wallace developed A.L.I.C.E. (Artificial Linguistic Internet Computer Entity) in the late 1990s. A.L.I.C.E., also known as Alicebot, is acknowledged for its pioneering programming using Artificial Intelligence Markup Language (AIML) which is an XML schema for creating natural language software agents. Wallace released the first version of AIML in July 2001. He has since established Pandorabots Inc. – an AI company that is distinguished as one of the world’s earliest chatbot hosting services and publishes the Pandora API on which A.L.I.C.E. was based and which enabled it to become a three-time Loebner winner in 2000, 2001, and 2004.
Jabberwacky:
Concurrent to A.L.I.C.E. developments, Jabberwacky was being conceived by British programmer, Rollo Carpenter. Jabberwacky was intended to simulate “natural human chat in an interesting, entertaining and humorous manner”. The emergence of the Internet provided Jabberwacky with a dynamic database of thousands of online human interactions from which to process responses. Jabberwacky under the guise of ‘George’ and ‘Joan’ won the Loebner Prize in 2005 and 2006 respectively.
Cleverbot:
In 2008, Jabberwacky launched a new iteration rebranded as Cleverbot. Like Jabberwacky, Cleverbot is designed to learn from its conversations with humans (more than 150 million to date according to Wikipedia).
cleverbot application. Cleverbot website (c) 2018. www.cleverbot.com
(accessed 3 March 2018)
It draws on past interactions to determine future questions and answers. To try out its capabilities, you can chat with Cleverbot on the official website: http://www.cleverbot.com/
IBM Watson:
In the endeavour to extend question answering (QA) capabilities posed in natural language, IBM Watson was conceived in 2006 as a QA computing system with the goal of outperforming human contestants on the U.S. TV game show Jeopardy! IBM Watson was developed as part of IBM’s DeepQA project by a research team led by David Ferrucci. Watson became the first computer to defeat contestants on the TV game show Jeopardy!, notably in a special match between Watson and Jeopardy! Champions, Ken Jennings and Brad Rutter. More recently, the Watson capabilities have evolved to take advantage of new deployment models (e.g. Watson on IBM Cloud) and new machine learning capabilities to “adapt and learn”.
Recent developments:
Chatbots in general are reaching milestones in artificial intelligence capability, as well as their pervasiveness in consumer facing products and services. For example in 2014, a chatterbot called Eugene Goostman won an AI contest marking the 60th anniversary of Turing’s death (Turing Test 2014 organised by the University of Reading) in which 33% of the event’s judges thought that Goostman was human. In development since 2001 and originating from St Petersburg, Goostman is portrayed as a 13-year-old Ukrainian boy. The Goostman bot has competed in several Loebner Prize contests since its creation, finishing second in 2005 and 2008.
Created from AIML technology by programmer, Steve Worswick, Mitsuku is a web-based chatbot available on Kik Messenger. It is among a growing number of sophisticated bots that can answer questions, play games, and capable of basic reasoning in QA. Mitsuku is a three-time winner of the Loebner prize in 2013, 2016, and recently in 2017.
Generally, the term ‘chatbot’ has referred to a software application that engages in a dialogue with a human using natural language. Most early advances have been associated with written language, but with advances in speech recognition, there is a narrowing of these associations. An early example is Naturally Speaking which was developed in 1975 by Dr James Baker (Carnegie Mellon University), as a simple speech understanding system that was called Dragon. Dr Baker worked on the system until 1982 when he and his wife, Dr Janet Baker, evolved the software into Dragon Systems, a Voice Recognition System. Other advances in the speech recognition sector have been made possible by VoiceXML which has been published in a series of standards since the release of version 1.0 in the year 2000. The World Wide Web Consortium (W3C) has recently announced a new W3C Community Group on Voice Interaction, which aims to explore beyond the system-initiated directed dialogs of VoiceXML applications.
In just a couple of years, there has been an exponential rise of voice assistants such as Apple Siri launched in 2010, Google Now in 2012, Amazon’s Alexa and Microsoft’s Cortana in 2015, and Google Assistant in 2016. Using natural language processing and Internet of Things (IoT) platforms, these assistants connect to web services to answer questions and respond to user requests. Recently, Google Home and Amazon Echo have started becoming consumer features in the home.
Social media platforms are similarly incorporating chatbot functionality. The year 2017 witnessed a particular hype cycle as Facebook opened up its Messenger platform and API to developers, providing an opportunity for anyone to build a simple chatbot on Facebook with ease. Twitter followed suite in April 2017 by opening its direct messaging channel to chatbots. WeChat has had chatbot functionalities for a few years now, and Slack and other messenger services are coming on board with open APIs.
Museums and chatbots:
Within this informal history of chatbot developments, it may not be surprising that museums and galleries have a track record in experimenting with new ways of communicating and with the use of emergent technologies to reach their audiences (3), (4), (5). Emerging free chatbot-creating platforms (e.g. Chatfuel, Chatterbot Eliza, among others) and the availability of open APIs, for instance, can offer both large and smaller museums the opportunity of experimenting with chatbots with relatively low effort while keeping costs and staff resources at a low level (5), (6).
Museums already have precedence in piloting technologies encompassing artificial intelligence and natural language processing with few resources (4), (5). Combining in-house or simple production methods with design thinking practices, museums can be enabled to develop interesting products.
There are in fact a growing number of Museums currently going down this route, and using bots to engage their audiences.
Below is a selected list of cultural organisations, discoverable at the time of this publication, that are experimenting with bots as part of their audience engagement programming:
Selected List of Museums using Chatbot Applications for Audience Engagement:
The Heinz Nixdorf MuseumsForum in Paderborn Germany has an early experience of using an avator bot introduced as MAX. Developed in 2004, MAX is a conversational agent that directly engages with visitors through a screen as a virtual museum guide.
The Cooper-Hewitt Smithsonian Design Museum, New York City:
The Cooper-Hewitt has been a pioneer in chatbot technologies with the creation of the Object Phone in 2013 – a service powered by Twillio that a visitor can text or call to ask for more information about a museum object in the collection. In 2016, Object Phone became a subscription service so that a visitor can receive a daily update. Anyone can participate in trialling the Object Phone: http://objectphone.cooperhewitt.org and signing up. In the words of Micah Walter, Director of Digital & Emerging Media:
“I think institutions like museums have a great opportunity in the chatbot space. If anything it represents a new way to broaden our reach and connect with people on the platforms they are already using. What’s more interesting to me is that chatbots themselves represent a way to interact with people that is by its very nature, bi-directional”.
Micah Walter, July 4, 2016.
Source: https://labs.cooperhewitt.org/2016/object-phone-the-continued-evolution-of-a-little-chatbot
San Francisco Museum of Modern Art (SFMOMA)
Send Me SFMOMA is an SMS service that provides an approachable, personal, and creative method of sharing the breadth of SFMOMA’s collection with the public. According to Jay Mollica, Creative Technologist, there are thousands of unseen artworks (only 5% are shown in Galleries at any one time) that can now be discovered through this application by texting 572-51 with the words “send me” followed by a keyword, a colour, or an emoji and the visitor will receive a related artwork image and caption via text message.
Source: www.sfmoma.org/read/send-me-sfmoma/ (accessed 3 March 2018)
Source: https://www.sfmoma.org/send-me-sfmoma/
Carnegie Museum of Art – Carnegie Museums, Pittsburgh.
Similar to the SFMOMA application, Carnegie Museum of Art has developed an SMS-based interaction called Muse which aims to leverage the Carnegie Studio’s user-centered design process and to make use of leading-edge technologies like natural language processing and image recognition. Source: Jeffrey Inscho, Carnegie Museum of Art, May 19, 2017. Source: https://studio.carnegiemuseums.org/introducing-muse-20a6f11c7c35
Anne Frank House in Amsterdam
In March 2017, Anne Frank House in Amsterdam launched its own Facebook Messenger chatbot that allows users to discover the History of Anne Frank – both her personal history and practical visitor information. Not simply a collections discovery bot, this application offers various conversation paths, allowing users to follow different paths in the Anne Frank story with concise information and links to additioal content, for example, excerpts from her diary to the context of World War II at the time. See: https://www.youtube.com/watch?v=YPH4vUWcN2U Anne Frank House – Published on Apr 3, 2017.
The National Art Museum of the Republic of Belarus
The National Art Museum in Belarus has also utilised the Facebook Messenger chatbot capability to create a conversational digital guide released in May 2017. A visitor can interact with the bot at:
https://chatfuel.com/bot/The-National-Art-Museum-of-the-Republic-of-Belarus-1904070043163955 . Source: The first Facebook chatbot guide for the National art museum of the Republic of Belarus available on Vimeo https://vimeo.com/213855387 – published by NODUCK, May 2017.
The Museum of Australian Democracy
The Museum marked its 50th anniversary of a landmark 1967 referendum in which Australians voted overwhelmingly to amend the Constitution to include Aboriginal people in the census and to allow the Commonwealth government to create laws for them. This landmark year lead to the launch of a referendum chatbot that allows visitors to learn about the historic and current impacts of this vote through chatting with it on Facebook Messenger. Directed towards children and accessible to adults, it uses simple gamification and responses, including emojis.
“Using a chatbot as a visitor engagement tool is innovative amongst Australia’s cultural institutions. It acts like history in your pocket and is helping MoAD spark a conversation about the significance of the 1967 referendum. We’re hoping it will be an effective way for people to get the facts, hear Indigenous perspectives on the referendum and reflect on its continuing relevance today.”
Source: Marni Pilgrim, Digital Engagement Manager. When a nation votes Yes, history is written and a chatbot is born. Canberra Times Media Release 25 May 2017
Twitterbots
Not unlike the quick adoption of Facebook Messenger among some of the Museum examples, there is a certain trend in Twitter bots such as the Museumbot that pulls open access images from a number of archives such as the Metropolitan Museum of Art. Other museum archive bots are steadily growing on Twitter, including the Los Angeles County Museum of Art LACMA bot, New York Public Library NYPL postcards bot, and the Museum of Modern Art MoMaR bot. A more comprehensive list can be found on Twitter from John Emerson.
The House Museums of Milan Chatbot : Case Study
This snapshot of museum and gallery innovations in chatbot deployment provides the background for a more detailed Case Study about the chatbot work undertaken for The House Museums of Milan (7). Of interest to readers is the twofold consideration of the developers and curators in the creation of The House Museum chatbot – namely:
- to attract and engage teenage audiences to these sites – a notoriously difficult public to engage in museums who are identified with high levels of distraction and highly adapted to the use of social media (6),(7),(8) AND
- to consider interactivity, such as gamification, that can interest and sustain attention by teenage and visitor audiences. (7),(8)
Gamification of chatbot applications is not yet pervasive among museum developments. The Referendum Bot developed for the Museum of Australian Democracy is one example of a gamified bot. The challenges encountered in the House project are described further below as an insider view. The process of chatbot development in this case was also intended to be model for how to make best use of a technology that is still in its infancy, and how to consider promising applications supporting audience interaction, for example, in self-guided tours and education (6), (9), (10).
Goals of the chatbot project:
The House Museums of Milan is a group of 4 historical homes in Milan (Poldi Pezzoli Museum, Bagatti Valsecchi Museum, Necchi Campiglio Villa and Boschi Di Stefano House Museum). When the House Museums launched a strategic initiative in 2016 that aimed to motivate people to visit the four museums through a single digital guide, they approached InvisibleStudio LTD (a London-based cultural innovation studio) to introduce gamification into the engagement process, specifically to attract younger audiences.
The founders of InvisibleStudio, Stefania Boiano and Giuliano Gaia, had already experimented with earlier chatbot technology in 2002, while working at the Museum of Science and Technology “Leonardo da Vinci” in Milan. You can find more details about this chatbot development in the MW 2003 paper: Make Your Museum Talk: Natural Language Interfaces For Cultural Institutions. (11)
This pioneering work provided important lessons in the development of this first chatbot. A key problem was that the chatbot was developed to mimic a Leonardo da Vinci character with whom the user would interact/”chat”. This set high expectations for the user experience, and led to frustrations when the bot was not able to understand the user beyond simple introductory chat. Consequently, user issues occurred quite soon in the conversation. (11)
With these lessons in hand, InvisibleStudio decided to change their approach when creating a chatbot for the House Museums project. The chatbot would only be used as a tool to help users solve a game set in the real physical environment of the museum. Thus, this approach could shift the user’s focus from the conversation with the chatbot to the actual exploration of the Museum galleries.
The chatbot game was developed using Facebook Messenger, and is directed mainly to young users and teenagers to engage them in exploring the four homes. Exploration is encouraged by users in looking for hidden clues that lead to a final discovery. The gaming activity is set in the context of fighting a mysterious Renaissance magician (based on a real historical figure) that provides a further incentive to engage with the application.
In this way users’ attention are drawn away from the limited conversational skills of the chatbot and invited to observe the collections with more attention while using the chatbot as a “virtual companion” in the game. Other key features had to be tweaked before publishing the application, such as making conversations more realistic by studying real chats on Facebook Messenger, referencing objects which the user can actually see “here and now” in the galleries, and finding the perfect length for the game (11), (12).
Another key challenge is the necessity of keeping open a continuous online connection between users and the chatbot. This can prove tricky in historical house museums, where the older infrastructure is comprised of complex layouts and thick walls which can prevent wifi or an even distribution of wifi connections.
The chatbot was tested with teenagers aged 16-18. The pilot was conducted by InvisibleStudio with 80 teenaged students from local high schools in Milan. This pilot provided the following results:
- 90% of students managed to complete the game
- 30% had connection problems
- 34% were worried for their data traffic
- 88% found the length of the game was right
- 72% evaluated the game as highly entertaining
- 66% found it a useful learning tool, especially if it was used with another student or in a small group.
These results offered some clear directions for the final development stages. Especially interesting for the developers was the fact that students liked using the chatbot in small groups, rather than on their own, because the game triggered collaboration within the team and created a friendly competitive environment with other teams. (12)
Challenges still need to be addressed. These are mainly related to the Facebook Messenger platform itself. For instance, teenagers seem to be abandoning it at an increasing rate. WhatsApp appears to be much more popular, but WhatsApp has not opened its API to third-party software yet. Thus creating a chatbot in WhatsApp is much more difficult at this stage although industry rumours suggest that WhatsApp will open its API soon (13). When this happens, museum chatbots could be developed for a potentially larger audience (e.g. teenagers), and potentially a larger uptake which does not depend on a subscription (e.g. Facebook) as the situation exists currently.
Concluding thoughts:
What emerged from this project is that the convergence of chatbots and gamification can be a powerful tool to involve younger, digital savvy generations visiting museums in novel and interesting ways (5), (6), (7), (9), (10). Our findings particularly suggest that users enjoy interacting with a chatbot in a game context, and this engagement can provide a smarter way of leading younger audiences to interact with objects and historic environments with greater attention. However with all the successes of this chatbot launch, there also remain challenges that need further consideration beyond the scope of this paper. As mentioned above, the availability of a wider range of chatbot platforms is one such challenge. A more involved issue is the pace and quality of the bot conversation which emerged as a critical aspect of this project. The present chatbot application required a bigger effort from the developing team to create engaging and realistic non-linear narrative lines, and this will be part of a continuing iteration in future developments.
Bibliography, References and Resources
(1) Dale, R. The return of the chatbots. Natural Language Engineering 22 (5) 2016 : 811–817. doi:10.1017/S1351324916000243
(2) Turing, A. M. (1950) Computing Machinery and Intelligence. Mind 49: 433-460.
(3) Borda, A. and Bowen, J. Smart Cities and Cultural Heritage: A Review of Developments and Future Opportunities. IN Proceedings. Electronic Visualisation and the Arts (EVA London 2017), BCS, London, pp 9-18.
(4) Bordoni, L., Mele, F. and Sorgente, A. (eds). Artificial Intelligence for Cultural Heritage. Newcastle: Cambridge Scholars, 2016.
(5) Boiano, S., Gaia, G. and Caldarini, M. Make Your Museum Talk: Natural Language Interfaces for Cultural Institutions. Museums and the Web 2003. URL: https://www.museumsandtheweb.com/mw2003/papers/gaia/gaia.html
(6) Boiano, S., Cuomo, P. and Gaia, G. Real-time Messaging Platforms for Storytelling and Gamification in Museums: A case history in Milan. IN Proceedings. Electronic Visualisation and the Arts (EVA 2016), London, UK, 12 – 14 July 2016. doi.org/10.14236/ewic/EVA2016.60. URL: http://ewic.bcs.org/content/ConWebDoc/56302
(7) Fors, V. Teenagers’ Multisensory Routes for Learning in the Museum. The Senses and Society, 8:3, 2016, 268-289, DOI: 10.2752/174589313X13712175020479
(8) Endo, Tasia. “Teens use tech to talk art: Amplifying teen voice and art interpretation.” MW2016: Museums and the Web 2016. Published March 10, 2016. Consulted September 29, 2017.
http://mw2016.museumsandtheweb.com/paper/teens-use-tech-to-talk-art-amplifying-teen-voice-and-art-interpretation/
(9) Kelly, L. and Russo, A. From Ladders of Participation to Networks of Participation: Social Media and the Museum Audiences. MW2008: Museum and the Web Conference, April 9–12, Montreal, Canada.
(10) Cawston, Rob, Daniel Efergan and Lindsey Green. “It’s in the game: can playful digital experiences help organisations connect with audiences in new ways?.” Museums and the Web: MW 2017. Published January 31, 2017. Consulted September 29, 2017.
https://mw17.mwconf.org/paper/its-in-the-game-can-playful-digital-experiences-help-organisations-connect-with-audiences-in-new-ways/
(11) Boiano S. and Gaia G., 3 Lessons learnt from Building our first Museum Chatbot… 15 years ago! Invisible Studio website. Consulted February 28th, 2018. URL: https://medium.com/@invisiblestudio/3-lessons-learnt-from-building-our-first-museum-chatbot-15-years-ago-7189c8a8fe6
(12) Boiano S. and Gaia G., 5 Tips for Involving Teenagers in Your Museum Using a Chatbot. Invisible Studio website. Consulted February 28th, 2018. URL: https://medium.com/@invisiblestudio/5-tips-for-involving-teenagers-in-your-museum-using-a-chatbot-bf88ff3ad568
(13) Mool, T. (2018) Chatbot Trends: The Year of the VoiceBot, WhatsApp Bots, MaaP. NativeMSG. Consulted February 28th, 2018 URL: https://nativemsg.com/blog/2018-chatbot-trends-the-year-of-the-voicebot-whatsapp-bots-maap