Language Detector Actor

This actor detects the language of each line of text using machine learning techniques. It allows users to input multiple lines of text, with each line representing a separate text sample. The actor will return the detected language for each line, along with confidence scores and alternative language guesses.

Features

• Language Detection: Automatically detects the language of each line of text.
• Confidence Scores: Provides the likelihood of the language detection for each guess.
• Alternative Guesses: In case the first guess isn't 100% accurate, it returns additional language options with probabilities.
• Multi-line Support: Accepts multiple text lines as input (one per line), with each line being processed individually.

Input

• Text: Paste or enter multiple lines of text (one text sample per line). Each line will be processed separately to detect the language.

Example Input

1Hello, how are you?
2Bonjour, comment ça va?
3これは日本語です。

Output

The actor will return the detected language for each line of text, along with alternative language guesses and their confidence scores.

Example Output

1[
2  {
3    "text": "Hello, how are you?",
4    "language": "en",
5    "confidence": 0.999995
6  },
7  {
8    "text": "Bonjour, comment ça va?",
9    "language": "fr",
10    "confidence": 0.999991
11  },
12  {
13    "text": "Hola, ¿cómo estás?",
14    "language": "es",
15    "confidence": 0.999995
16  },
17  {
18    "text": "これは日本語です。",
19    "language": "ja",
20    "confidence": 1.0
21  }
22]

How It Works

1. Input Processing: Users input multiple lines of text (one line per text sample). Each line is parsed and passed to the language detection model.
2. Language Detection: The actor uses a machine learning model (Naive Bayes classifier) to detect the language based on n-gram patterns found in the input.
3. Output: For each line of text, the actor returns the detected language, along with the confidence score and alternative guesses.

Deployment

This actor is deployed on Apify and can be used via the Apify Console. Once deployed, users can provide text as input and retrieve language detection results through the UI.

Usage

1. Go to the Apify Console.
2. Run the actor with your text input.
3. Download the results or view them in the Apify UI.

Limitations

• The actor is best suited for shorter texts, such as sentences or short paragraphs.
• Accuracy can vary with very short or ambiguous text inputs.