Conversation practice is essential for second-language acquisition and necessary for learners to reach an acceptable communicative level. In an ideal scenario, students should regularly hold conversations with native speakers of the target language, but this is often not possible. Although the teachers use the target language during the classes, they cannot offer a continuous conversation with each student, and they are usually not available wherever and whenever the student requires them to practice. This chapter presents the development and use of a mobile application to hold conversations in the Japanese language. The objective is to provide a software tool to improve the level of communicative competence (both inside and outside a classroom environment). So, the authors created and used a conversational agent (chatbot) using Dialogflow (a Google API), which is connected to the application's interface through the Internet using a client access token to give responses to user inputs in real time.
TopBackground
It is estimated that around 7,000 different languages are spoken worldwide. 90% of these are spoken by less than 100,000 people, such as those from tribes or populations with different dialects. From 150 to 200 languages are spoken by one million people, while 46 types of language are spoken by only one individual (BBC, 2014).
With around 128 million speakers, of which 127 million are native, Japan occupies the eleventh place worldwide. As you can see, it is a language that is around 99% concentrated in its country of origin, but whose importance remains in the fact that it is currently positioned as the third world power and is expected to continue in that position in the following years (Portillo, 2019).
Due to the globalization that has been generated in recent decades and the need for communication that this represents, technologies have been developed to facilitate the learning of a foreign language or that directly facilitate communication between people who do not speak the same language. Although chatbots (or chatterbots) have been used for different purposes for several years, little has been said about their participation as assistants for the practice of a foreign language, especially through the use of inputs and outputs using voice.
In the work of Angga, Fachri, Angga, Suryadi, and Agushinta (2015), it is mentioned that text inputs generated by users to interact with chatbots are considered inefficient because the conversation becomes unnatural and uninteresting for the user. Based on this, the author proposed the use of a voice interface, which uses technologies such as the recognition and generation of spoken language to obtain the inputs and generate the appropriate outputs. A third-dimensional avatar was integrated, capable of generating different facial expressions in real-time to give the user experience closer to a conversation with a human being. The author proposes the use of webcams in future designs as an additional tool to analyze the emotions and reactions of the user. It should be noted that this design was focused on the customer service area and not on teaching.