Just enter any text in a supported language (Arabic, Chinese, Danish, English, French, German, Russian or Spanish) or select an example from the dropdown menu and press the Submit button to identify the language and render the corresponding Semantic Fingerprint. You can explore the meanings associated with different parts of the Fingerprint by moving your mouse over its active regions, shown in blue.

Use Cases

  • Text Preprocessing
  • Content Classification
  • Text Filtering

How It Works

The language of the input text is determined based on profiles for supported languages. Distinguishing features of each language are extracted and used with a Bayesian filter to detect the language of the input text. The most appropriate encoding Retina available for the identified language is then returned and used to generate a Semantic Fingerprint for the input text.

Build Your Own Application

Language detection is an upcoming feature soon to be available as part of the API. Once available, it will facilitate further text processing by allowing the user to select an appropriate encoding Retina from a list of supported languages. This demo also makes use of an another API call to retrieve a visual rendering of the text's Semantic Fingerprint.

  • Language Detection Endpoint (coming soon) - Return the Retina corresponding to the language of the input text.
  • Text Endpoint - Compute the semantic fingerprint for an input text.

You can begin integrating's services into your own application after registering for your own free API key.