OpenNLP Sentence Detection for Android Example

Using OpenNLP for Sentence Detection in Android example.

In natural language processing, deciding where is starting and ending points of a sentence in a paragraph is very important.

Here I use OpenNLP to extract the position of sentences in a paragraph. Sentence Detection depends on the language in the paragraph. In this example, I extracted sentence in English. We can change the model for other languages by downloading it here

Download and put the .bin file in asset folder.

First, Adding the following code in your app/build.gradle file to add openNLP library:

    implementation group: 'org.apache.opennlp', name: 'opennlp-tools', version: '1.9.3'

Second, we need to load the model from asset folder

        //Loading sentence detector model
        InputStream inputStream = null;
        SentenceModel model = null;
        
        try {
        
            inputStream = getAssets().open("en-sent.bin");
            model = new SentenceModel(inputStream);

        } catch (IOException e) {
            e.printStackTrace();
        }

Finally, we instantiating the SentenceDetectorME class and detect sentences from input text.

        //Instantiating the SentenceDetectorME class
        SentenceDetectorME detector = new SentenceDetectorME(model);
        
        //Detecting the sentence
        sentences = detector.sentDetect(inputEt.getText().toString());
        sentencesIndex = detector.sentPosDetect(inputEt.getText().toString());
        }

Screenshot

Copy the paragraph and paste in the InputText:

Click EXECUTE button and see result: