Europeana Multilingual Chatbot proves to be another innovative intersection of combining language, culture and technology, it came to light to create a conversational interface designed to facilitate and enrich users’ navigation of cultural heritage sites.
Conversational interfaces, or chatbots, are increasingly becoming a popular interface to interact with users in e- commerce, online banking, transportation, healthcare and other industries. They are a means of website users to obtain detailed answers to possible queries, inspiration for purchases, or for communicating with multiple brands. Chatbots use natural language to converse with users, understand their needs and provide them with this information and other human-like services.
The Culture Chatbot - Europeana Multilingual Chatbot project aims to exploit the potential of conversational interfaces for users that are engaging with cultural heritage content. It is a standalone chatbot platform that will be interacting with online users on a variety of online platforms - websites, social media, instant messaging tools - providing an easy and novel access to cultural heritage content, coming from Europeana, its aggregators as well as other cultural heritage institutions. The chatbot will guide the website navegators, using an intuitive natural-language interface, across Europeana collections. This will allow them to discover new cultural heritage resources based on their expressed interests or by providing inspiration on what type of material to search for.
The project partners- Culture Italia, Polin and Amsterdam’s Jewish Historical Museum- constitute an array of cultural heritage organizations that will use the chatbot for their websites. In the scope of this project, the chatbot will be available in English, Italian, Dutch, Spanish and Polish- with potential of adding more languages.
In the meeting, economic and technical frameworks for the project were discussed. After establishing the project’s economic framework, technical issues with traditional online experiences were reviewed, and how a chatbot can aid this. Moreover, it was established that the chatbot will comprise a hybrid between a rule-based chatbots, which are designed to respond to specific rules or patterns with predetermined answers; and an AI chatbot, which by using sophisticated cognitive and natural language processing capabilities, can understand requests but also context, intent, emotion. This hybrid model was chosen as it is believed to enrich the service user’s experience when navigating the cultural sites.
Following, the Europeana Multilingual Chatbot's architecture was established, including its annotation tools, the channels it will be used in, OS and its framework.