File upload
Text files: Pinpoint automatically detects the language of written text in non-audio files. Non-audio files can contain a mix of languages.
Audio files: Make sure that Pinpoint knows the language of an audio file before you upload it by configuring your audio file language. If an audio file contains multiple languages, Pinpoint can extract only words in the language specified for audio file uploads. If you want to upload audio files in multiple languages, you'll need to change your audio file language setting before uploading each file.
Cross-language search
Pinpoint can search for strings and entities in the language that you use in your search term, and can also find similar terms in other. For example, searching a Spanish language document for the string "New Hampshire" will also find the string "Nuevo Hampshire" in the Spanish language document.
Entity display language
Choose the display language for document entities. This setting affects the display language for entities, but does not search or filtering behavior. So, for example, if you choose Spanish as the entity display language, filtering on the entity "Nuevo Hampshire" can return results for "New Hampshire" in all languages in your documents.
Entity display language is configured at the account level, and applies to any collection that you view. Changing this setting does not affect other users, and requires only view-only permission.
To change your entity display language:
- From the Documents view, click settings
.
- Under Entity display language select a language.
- Click Save changes. Changes will apply only to you, but in all collections.
.Audio file spoken language
In order for Pinpoint to transcribe audio text properly, Pinpoint needs to know the audio file language before you upload it. Any words in a language other than the audio upload language will be improperly transcribed. The default audio language is based on your preferred language for your Google account.
Pinpoint can transcribe audio in the following languages: English (US, UK or AU dialects), Czech, Dutch, French, German, Hebrew, Indonesian, Italian, Polish, Portuguese, Russian, Ukrainian, and Turkish.
Note that changing your audio file language has no effect on files already uploaded to your collection.
Audio file upload language is configured per-user across all collections. Changing this setting will not affect other users.
To change your audio file upload language:
- From the Documents view, Click settings
- Under Audio files spoken language select the language of your recordings. Note that this setting applies to all your collections.
- Click Save changes.
- Upload your audio files. (You can upload a mix of audio and non-audio files.)
- After processing, each audio file will be published as a PDF transcript in your collection, with an embedded player for the original audio. These transcripts are searchable the same as all other documents in your collection.
If you provide the incorrect language for an audio file, you will need to update your audio file language setting and re-upload the file.