SwedenLivingWorld world_2_fill WorldBusiness BusinessSports sports-soccer SportsEntertainmentEntertain

Old dialects are gold worth for AI

How can some Halland guys make AI technology better? When American speech-to-text models are to learn Swedish, old dialect recordings have proven to be a goldmine.

» Published: October 12 2024

Old dialects are gold worth for AI
Photo: Uppsala universitet

At the Institute for Language and Folk Culture (Isof), they have worked extensively with dialect collection, primarily between 1935 and 1970, when researchers traveled around the country recording how people spoke.

The initial purpose was largely to preserve traditional dialects, explains Annette C Torensjö, head of the Department of Archives and Research in Uppsala at Isof.

The recordings were used for many years mainly by researchers, but last year, the KB-Lab at the Royal Library got in touch. Leonora Vesterbacka is a senior data scientist and leads a project to train speech-to-text models, which "translate" speech into written text.

A Small Part of Swedish

American AI models are often trained on massive multilingualism, and then the Swedish part becomes very small.

There is a model from Open AI called Whisper, which is trained on 680,000 hours of various languages, of which around 2,000 hours are Swedish. That's 0.3 percent Swedish, explains Leonora Vesterbacka.

The models work well on standard Swedish, but worse on dialects. That's where the Halland dialects, and others who had their dialects recorded, come in.

If someone had told me in the 1980s that the new gold would be an old weird roll of recordings in an archive, I would have thought they were joking.

They also use protocols and recordings from the Riksdag Administration.

They are responsible for recording what is said in the Riksdag and then making it available and releasing protocols. This has been going on for a long time.

In Sweden, we are so good at preserving everything. It's amazing to see that it can be used in the future as well.

Making Available

When the models are fully trained, they can, for example, be used to transcribe medical records and meetings or write subtitles for TV broadcasts. They can also be used to make spoken material, such as podcasts and TV broadcasts from authorities, available.

At Isof, they had not believed that their old dialect recordings would become an important step into the future.

I'm very pleased that our dialect recordings really have such great relevance now, that this is something that can be worked with as part of social development, says Annette C Torensjö.

A dialect is a language variant spoken by the inhabitants of a geographically defined area. It differs from the standard language as well as from adjacent dialects. The differences can concern language features at all levels: phonological, morphological, lexical, and syntactic.

Sociolect, or social dialect, is a language variant that is characteristic of a particular social group. Sociolects are often limited to pronunciation, grammar, and vocabulary, and are also regionally restricted, so that members of the same social class in different parts of a language area vary from each other.

Source: ne.se

Tags

TTT
By TTThis article has been altered and translated by Sweden Herald

More news

Centre Party: There are cheaper alternatives to new nuclear power

Centre Party: There are cheaper alternatives to new nuclear power

Woman assaulted in Märsta – man charged with murder

Woman assaulted in Märsta – man charged with murder

Sick parrot didn't get care – owner prosecuted

Sick parrot didn't get care – owner prosecuted

So, public servants' protection will be strengthened

So, public servants' protection will be strengthened

The Gaming Inspectorate Criticizes S Lotteries

The Gaming Inspectorate Criticizes S Lotteries

Record number flee climate catastrophes

Record number flee climate catastrophes

Hultsfred introduces ban on laughing gas

Hultsfred introduces ban on laughing gas

Complaints to the Parliamentary Ombudsman increase – "there are risks"

Complaints to the Parliamentary Ombudsman increase – "there are risks"

This is how you get a grip on the future shoreline

This is how you get a grip on the future shoreline

Nine Convicted of Fraud via Dating Sites

Nine Convicted of Fraud via Dating Sites

Woman found dead – man arrested for murder

Woman found dead – man arrested for murder

Investigation of Millennium Failure Completed

Investigation of Millennium Failure Completed

16-year-old convicted of raping a minor girl

16-year-old convicted of raping a minor girl

Threatening S-Politician Relieved of Teaching Job

Threatening S-Politician Relieved of Teaching Job

Two Sentenced to 16 Years for Murder at Nightclub

Two Sentenced to 16 Years for Murder at Nightclub

Investigation: Violence and Threats Against Witnesses in the Lundin Case

Investigation: Violence and Threats Against Witnesses in the Lundin Case

That's why the Constitutional Committee is pausing its investigation of the Prime Minister

That's why the Constitutional Committee is pausing its investigation of the Prime Minister

All newborns to receive medicine against RS virus

All newborns to receive medicine against RS virus

School leader convicted - demanded nude photos from student

School leader convicted - demanded nude photos from student

The Minister: ADHD diagnoses are sold for payment

The Minister: ADHD diagnoses are sold for payment

Indictment: Shot and killed two in an apartment in Norrköping

Indictment: Shot and killed two in an apartment in Norrköping

Dog in collision with police car – being re-examined

Dog in collision with police car – being re-examined

The Gold Find – "Sensational" Neck Ring

The Gold Find – "Sensational" Neck Ring

Öresund Bridge reopened after car fire

Öresund Bridge reopened after car fire

Teacher Fired – Accused of Taping Children's Mouths

Teacher Fired – Accused of Taping Children's Mouths

Hospital evacuated after leak – now rectified

Hospital evacuated after leak – now rectified

Boy at foster home posted murder ads

Boy at foster home posted murder ads

The Investigation into the Oceana Fire Completed

The Investigation into the Oceana Fire Completed

Man charged with explosion and shot horse

Man charged with explosion and shot horse

Police Officer Charged - Suspected of Leaking Confidential Information

Police Officer Charged - Suspected of Leaking Confidential Information

Municipalities reject uranium mining: "Unacceptable"

Municipalities reject uranium mining: "Unacceptable"

Spring warmth for the weekend – is it here to stay?

Spring warmth for the weekend – is it here to stay?

Fast pace in this year's lynx hunt

Fast pace in this year's lynx hunt

The Swedish Association of Health Professionals Slams Higher Patient Fee: "Devastating"

The Swedish Association of Health Professionals Slams Higher Patient Fee: "Devastating"

Burnout syndrome to be removed: "I think it will get better"

Burnout syndrome to be removed: "I think it will get better"

Forced to Leave Political Assignments After Threats

Forced to Leave Political Assignments After Threats

Doctor Charged with Causing Patient's Death

Doctor Charged with Causing Patient's Death

Crime victims to receive damages more easily

Crime victims to receive damages more easily

Furious over more expensive medicines

Furious over more expensive medicines

Researchers: "Stop the digital national tests"

Researchers: "Stop the digital national tests"