Commit c05629c5 authored by Manuel Segimon's avatar Manuel Segimon
Browse files

Refactor Corrector class to handle original sentence in top 5

parent 2bf2c37d

src/main/java/edu/bu/LanguageCorrection/Corrector.java

+21 −1
Original line number Diff line number Diff line
@@ -105,8 +105,24 @@ public class Corrector { 
            }

        }



        // Check if the original sentence is in the top 5

        boolean originalInTop5 = false;

        for (String sentence : topSentences) {

            sentence = sentence.split(" \\| Score: ")[0]; // Remove score

            System.out.println(sentence);

            if (sentence != null && sentence.equals(inputSentence)) {

                originalInTop5 = true;

                break;

            }

        }



        if (originalInTop5) {

            String[] result = new String[0];

            return result;

        } else {

            return topSentences;

        }

    }



    // Helper class to manage sentences and their scores

    class SentenceScorePair implements Comparable<SentenceScorePair> {
@@ -192,6 +208,10 @@ public class Corrector { 
                for (String sentence : sentences) {

                    sentence = sentence.replaceAll("[^a-zA-Z0-9\\s]", "");

                    String[] corrected = corrector.correct(sentence);

                    if (corrected.length == 0) {

                        System.out.println(sentence + " | No corrections needed.");

                        continue;

                    }

                    System.out.println(sentence + " | Corrected Sentence Suggestions:");

                    printSentencesInOrderOfChanges(corrected, sentence);

                }

src/main/java/edu/bu/LanguageCorrection/MainApp.java

+8 −0
Original line number Diff line number Diff line
@@ -273,8 +273,13 @@ public class MainApp extends JFrame { 
            for (String sentence : sentences) {

                sentence = sentence.replaceAll("[^a-zA-Z0-9\\s]", "");

                String[] corrected = corrector.correct(sentence);

                if (corrected.length == 0) {

                    result.append(sentence + " | No corrections found for this sentence.\n\n");

                    continue;

                }

                result.append(sentence + " | Corrected Sentence Suggestions:\n");

                printSentencesInOrderOfChanges(corrected, sentence, result);

                result.append("\n");

            }



            resultArea.setText(result.toString());
@@ -286,6 +291,9 @@ public class MainApp extends JFrame { 
                String sentenceNew = sentence.replaceAll("[^a-zA-Z0-9\\s]", "");

                String[] corrected = corrector.correct(sentenceNew);

                int correctedLength = corrected.length;

                if (correctedLength == 0) {

                    continue;

                }

                for (String s : corrected) {

                    if (s == null || s.isEmpty()) {

                        correctedLength--;

src/main/java/edu/bu/LanguageCorrection/TrieNode.java

+9 −2
Original line number Diff line number Diff line
@@ -50,6 +50,13 @@ public class TrieNode implements Serializable, Cloneable { 
        return (float) node.childCounts / node.children.size();

    }



    private int factorial(int n) {

        int res = 1, i;

        for (i = 2; i <= n; i++)

            res *= i;

        return res;

    }



    public float perplexity(String phrase) {

        float logProb = 0;

        String[] words = phrase.split(" ");
@@ -63,8 +70,8 @@ public class TrieNode implements Serializable, Cloneable { 
            logProb += Math.log((probability(currentPhrase)));

        }

        float perplexity = (float) Math.pow(2, -logProb);

        System.out.println("Perplexity of phrase (" + phrase + ") : " + perplexity);

        return perplexity;

        //System.out.println("Perplexity of phrase (" + phrase + ") : " + ((float) perplexity / factorial(words.length)));

        return (float) perplexity / factorial(words.length);

    }

    

    public byte[] serialize() {