@Override
public AggregateResult doMeasureUser(TestUser user, MeanAccumulator context) {
List<ScoredId> recommendations = user.getRecommendations(30, candidates, exclude);
if (recommendations == null || recommendations.isEmpty()) {
context1.add(0.0);
context5.add(0.0);
context10.add(0.0);
context15.add(0.0);
context20.add(0.0);
context25.add(0.0);
context30.add(0.0);
return null;
}
double ndcg1 = measureUser(user, context1, recommendations, 1);
double ndcg5 = measureUser(user, context5, recommendations, 5);
double ndcg10 = measureUser(user, context10, recommendations, 10);
double ndcg15 = measureUser(user, context15, recommendations, 15);
double ndcg20 = measureUser(user, context20, recommendations, 20);
double ndcg25 = measureUser(user, context25, recommendations, 25);
double ndcg30 = measureUser(user, context30, recommendations, 30);
return new AggregateResult(ndcg1, ndcg5, ndcg10, ndcg15, ndcg20, ndcg25, ndcg30);
}
public double measureUser(
TestUser user, MeanAccumulator context, List<ScoredId> recommendations, int listSize) {
if (recommendations.size() > listSize) {
recommendations = new ArrayList<ScoredId>(recommendations.subList(0, listSize));
}
SparseVector ratings = user.getTestRatings();
LongList ideal = ratings.keysByValue(true);
if (ideal.size() > listSize) {
ideal = ideal.subList(0, listSize);
}
double idealGain = computeDCG(ideal, ratings);
LongList actual = new LongArrayList(recommendations.size());
for (ScoredId id : recommendations) {
actual.add(id.getId());
}
double gain = computeDCG(actual, ratings);
double score = gain / idealGain;
context.add(score);
return score;
}