"Hirosaki University x AI x Tsugaru dialect project" is currently calling for the provision of Tsugaru dialect voice data.

The project started in 2017 (Heisei 29). Using artificial intelligence (AI) technology, we aim to develop software that converts Tsugaru dialect voice into standard language text and automatically translates it. It is expected to be useful in call centers and medical sites where Tsugaru dialect may be a problem.

In November 2020, the "Tsugaru Language Dictionary" was opened to the public, which can index about 10,000 Tsugaru dialect example data collected through open recruitment. Professor Masayuki Imai of the Graduate School of Science and Engineering, who is involved in the project, said, "It is generally said that about 200,000 data are required for automatic language translation. The data is still insufficient."

The voice data recruited this time can be posted from the project's homepage, and according to Mr. Imai, any trivial voice will do. "We are looking for Tsugaru dialects regardless of age or gender. For example, you can use the recording function on your smartphone to send recorded data. I want you to use just one word, not everyday conversation," says Mr. Imai. "We are also developing an app that will allow you to easily post in the near future."

The recruitment is open until March 31, but the recruitment will continue with the example data. “We will scrutinize it once, convert it into text, and add it as data to improve accuracy,” says Imai.

Mr. Imai is from outside the prefecture and may have trouble communicating with local residents living in the neighborhood. "The meaning of Tsugaru dialect may differ depending on the pronunciation and intonation of the same word. By collecting voice data, we can approach the creation of a system that anyone can understand, which is the final goal of Tsugaru dialect," he said.