Examples with JPlagResult de.jplag.JPlagResult used on opensource projects

Example 1 with JPlagResult

use of de.jplag.JPlagResult in project ArTEMiS by ls1intum.

the class TextPlagiarismDetectionService method checkPlagiarism.

/**
 * Download all submissions of the exercise, run JPlag, and return the result
 *
 * @param textExercise        to detect plagiarism for
 * @param similarityThreshold ignore comparisons whose similarity is below this threshold (%)
 * @param minimumScore        consider only submissions whose score is greater or equal to this value
 * @param minimumSize         consider only submissions whose size is greater or equal to this value
 * @return a zip file that can be returned to the client
 * @throws ExitException is thrown if JPlag exits unexpectedly
 */
public TextPlagiarismResult checkPlagiarism(TextExercise textExercise, float similarityThreshold, int minimumScore, int minimumSize) throws ExitException {
    long start = System.nanoTime();
    String topic = plagiarismWebsocketService.getTextExercisePlagiarismCheckTopic(textExercise.getId());
    // TODO: why do we have such a strange folder name?
    final var submissionsFolderName = "./tmp/submissions";
    final var submissionFolderFile = new File(submissionsFolderName);
    submissionFolderFile.mkdirs();
    final List<TextSubmission> textSubmissions = textSubmissionsForComparison(textExercise, minimumScore, minimumSize);
    final var submissionsSize = textSubmissions.size();
    log.info("Save text submissions for JPlag text comparison with {} submissions", submissionsSize);
    if (textSubmissions.size() < 2) {
        log.info("Insufficient amount of submissions for plagiarism detection. Return empty result.");
        TextPlagiarismResult textPlagiarismResult = new TextPlagiarismResult();
        textPlagiarismResult.setExercise(textExercise);
        textPlagiarismResult.setSimilarityDistribution(new int[0]);
        return textPlagiarismResult;
    }
    AtomicInteger processedSubmissionCount = new AtomicInteger(1);
    textSubmissions.forEach(submission -> {
        var progressMessage = "Getting submission: " + processedSubmissionCount + "/" + textSubmissions.size();
        plagiarismWebsocketService.notifyInstructorAboutPlagiarismState(topic, PlagiarismCheckState.RUNNING, List.of(progressMessage));
        submission.setResults(new ArrayList<>());
        StudentParticipation participation = (StudentParticipation) submission.getParticipation();
        participation.setExercise(null);
        participation.setSubmissions(null);
        String participantIdentifier = participation.getParticipantIdentifier();
        if (participantIdentifier == null) {
            participantIdentifier = "unknown";
        }
        try {
            textSubmissionExportService.saveSubmissionToFile(submission, participantIdentifier, submissionsFolderName);
        } catch (IOException e) {
            log.error(e.getMessage());
        }
        processedSubmissionCount.getAndIncrement();
    });
    log.info("Saving text submissions done");
    JPlagOptions options = new JPlagOptions(submissionsFolderName, LanguageOption.TEXT);
    options.setMinimumTokenMatch(minimumSize);
    // Important: for large courses with more than 1000 students, we might get more than one million results and 10 million files in the file system due to many 0% results,
    // therefore we limit the results to at least 50% or 0.5 similarity, the passed threshold is between 0 and 100%
    options.setSimilarityThreshold(similarityThreshold);
    log.info("Start JPlag Text comparison");
    JPlag jplag = new JPlag(options);
    JPlagResult jPlagResult = jplag.run();
    log.info("JPlag Text comparison finished with {} comparisons. Will limit the number of comparisons to 500", jPlagResult.getComparisons().size());
    log.info("Delete submission folder");
    if (submissionFolderFile.exists()) {
        FileSystemUtils.deleteRecursively(submissionFolderFile);
    }
    TextPlagiarismResult textPlagiarismResult = new TextPlagiarismResult();
    textPlagiarismResult.convertJPlagResult(jPlagResult);
    textPlagiarismResult.setExercise(textExercise);
    log.info("JPlag text comparison for {} submissions done in {}", submissionsSize, TimeLogUtil.formatDurationFrom(start));
    plagiarismWebsocketService.notifyInstructorAboutPlagiarismState(topic, PlagiarismCheckState.COMPLETED, List.of());
    return textPlagiarismResult;
}

Example 2 with JPlagResult

use of de.jplag.JPlagResult in project Artemis by ls1intum.

the class TextPlagiarismDetectionService method checkPlagiarism.

/**
 * Download all submissions of the exercise, run JPlag, and return the result
 *
 * @param textExercise        to detect plagiarism for
 * @param similarityThreshold ignore comparisons whose similarity is below this threshold (%)
 * @param minimumScore        consider only submissions whose score is greater or equal to this value
 * @param minimumSize         consider only submissions whose size is greater or equal to this value
 * @return a zip file that can be returned to the client
 * @throws ExitException is thrown if JPlag exits unexpectedly
 */
public TextPlagiarismResult checkPlagiarism(TextExercise textExercise, float similarityThreshold, int minimumScore, int minimumSize) throws ExitException {
    long start = System.nanoTime();
    String topic = plagiarismWebsocketService.getTextExercisePlagiarismCheckTopic(textExercise.getId());
    // TODO: why do we have such a strange folder name?
    final var submissionsFolderName = "./tmp/submissions";
    final var submissionFolderFile = new File(submissionsFolderName);
    submissionFolderFile.mkdirs();
    final List<TextSubmission> textSubmissions = textSubmissionsForComparison(textExercise, minimumScore, minimumSize);
    final var submissionsSize = textSubmissions.size();
    log.info("Save text submissions for JPlag text comparison with {} submissions", submissionsSize);
    if (textSubmissions.size() < 2) {
        log.info("Insufficient amount of submissions for plagiarism detection. Return empty result.");
        TextPlagiarismResult textPlagiarismResult = new TextPlagiarismResult();
        textPlagiarismResult.setExercise(textExercise);
        textPlagiarismResult.setSimilarityDistribution(new int[0]);
        return textPlagiarismResult;
    }
    AtomicInteger processedSubmissionCount = new AtomicInteger(1);
    textSubmissions.forEach(submission -> {
        var progressMessage = "Getting submission: " + processedSubmissionCount + "/" + textSubmissions.size();
        plagiarismWebsocketService.notifyInstructorAboutPlagiarismState(topic, PlagiarismCheckState.RUNNING, List.of(progressMessage));
        submission.setResults(new ArrayList<>());
        StudentParticipation participation = (StudentParticipation) submission.getParticipation();
        participation.setExercise(null);
        participation.setSubmissions(null);
        String participantIdentifier = participation.getParticipantIdentifier();
        if (participantIdentifier == null) {
            participantIdentifier = "unknown";
        }
        try {
            textSubmissionExportService.saveSubmissionToFile(submission, participantIdentifier, submissionsFolderName);
        } catch (IOException e) {
            log.error(e.getMessage());
        }
        processedSubmissionCount.getAndIncrement();
    });
    log.info("Saving text submissions done");
    JPlagOptions options = new JPlagOptions(submissionsFolderName, LanguageOption.TEXT);
    options.setMinimumTokenMatch(minimumSize);
    // Important: for large courses with more than 1000 students, we might get more than one million results and 10 million files in the file system due to many 0% results,
    // therefore we limit the results to at least 50% or 0.5 similarity, the passed threshold is between 0 and 100%
    options.setSimilarityThreshold(similarityThreshold);
    log.info("Start JPlag Text comparison");
    JPlag jplag = new JPlag(options);
    JPlagResult jPlagResult = jplag.run();
    log.info("JPlag Text comparison finished with {} comparisons. Will limit the number of comparisons to 500", jPlagResult.getComparisons().size());
    log.info("Delete submission folder");
    if (submissionFolderFile.exists()) {
        FileSystemUtils.deleteRecursively(submissionFolderFile);
    }
    TextPlagiarismResult textPlagiarismResult = new TextPlagiarismResult();
    textPlagiarismResult.convertJPlagResult(jPlagResult);
    textPlagiarismResult.setExercise(textExercise);
    log.info("JPlag text comparison for {} submissions done in {}", submissionsSize, TimeLogUtil.formatDurationFrom(start));
    plagiarismWebsocketService.notifyInstructorAboutPlagiarismState(topic, PlagiarismCheckState.COMPLETED, List.of());
    return textPlagiarismResult;
}

Example 3 with JPlagResult

use of de.jplag.JPlagResult in project JPlag by jplag.

the class NormalComparisonStrategy method compareSubmissions.

@Override
public JPlagResult compareSubmissions(SubmissionSet submissionSet) {
    boolean withBaseCode = submissionSet.hasBaseCode();
    if (withBaseCode) {
        compareSubmissionsToBaseCode(submissionSet);
    }
    List<Submission> submissions = submissionSet.getSubmissions();
    long timeBeforeStartInMillis = System.currentTimeMillis();
    int i, j, numberOfSubmissions = submissions.size();
    Submission first, second;
    List<JPlagComparison> comparisons = new ArrayList<>();
    for (i = 0; i < (numberOfSubmissions - 1); i++) {
        first = submissions.get(i);
        if (first.getTokenList() == null) {
            continue;
        }
        for (j = (i + 1); j < numberOfSubmissions; j++) {
            second = submissions.get(j);
            if (second.getTokenList() == null) {
                continue;
            }
            compareSubmissions(first, second, withBaseCode).ifPresent(it -> comparisons.add(it));
        }
    }
    long durationInMillis = System.currentTimeMillis() - timeBeforeStartInMillis;
    return new JPlagResult(comparisons, submissionSet, durationInMillis, options);
}