Peer-reviewed research

Uso de la entonación para identificar cuándo usar la tilde diacrítica en el reconocimiento automático del habla (Use of intonation to identify when to use the diacritical accent mark in automatic speech recognition)

Constantino Bolaños Araya et al., Kañina, Universidad de Costa Rica, 2016

Read on October 17, 2026

What it is: A Costa Rican journal article, in Spanish with an English title, that proposes using intonation to decide when a Spanish recognizer should write the diacritical tilde. It opens by describing missing tildes as one of the most common errors in Spanish automatic transcription.
How much weight it carries: Peer-reviewed university journal article. It is a proposal paper with its own small tests, not a large benchmark, and it describes recognizers of its time (2016), so current engines may do better. Use it for the diagnosis of the error, not for a rate of how often it happens.

Facts taken from this source

The authors say that in Spanish, one of the most common errors of automatic speech recognition is the omission of the diacritical tilde.
“uno de los errores más comunes de esta tecnología, es la omisión de la tilde diacrítica” (Introduction (Resumen and opening paragraph))
The authors attribute the missing tildes mainly to speech recognition techniques that ignore stress, the stressed syllable of a word.
“las técnicas utilizadas en el reconocimiento automático del habla ignoran el acento, es decir, la sílaba acentuada de una palabra” (Introduction (opening paragraph))

Peer-reviewed research

Constantino Bolaños Araya et al., Kañina, Universidad de Costa Rica, 2016

Read on October 17, 2026

What it is: A Costa Rican journal article, in Spanish with an English title, that proposes using intonation to decide when a Spanish recognizer should write the diacritical tilde. It opens by describing missing tildes as one of the most common errors in Spanish automatic transcription.
How much weight it carries: Peer-reviewed university journal article. It is a proposal paper with its own small tests, not a large benchmark, and it describes recognizers of its time (2016), so current engines may do better. Use it for the diagnosis of the error, not for a rate of how often it happens.

The authors say that in Spanish, one of the most common errors of automatic speech recognition is the omission of the diacritical tilde.
“uno de los errores más comunes de esta tecnología, es la omisión de la tilde diacrítica” (Introduction (Resumen and opening paragraph))
The authors attribute the missing tildes mainly to speech recognition techniques that ignore stress, the stressed syllable of a word.
“las técnicas utilizadas en el reconocimiento automático del habla ignoran el acento, es decir, la sílaba acentuada de una palabra” (Introduction (opening paragraph))