private static void showNeighboursAndSimilaritiesOfUserSeven(Dictionary <int, User> allUsers)
{
var userSeven = allUsers[7];
ISimilarity similarityMeasure = new EuclideanSimilarity();
var userSevenNeighboursAndSimilarities = userSeven.getNearestNeighboursAndSimilarities(3, allUsers.Values.ToList(), similarityMeasure);
Console.WriteLine("");
Console.WriteLine("Neighbours and similarities using EUCLIDEAN: ");
outputNeighbours(userSevenNeighboursAndSimilarities);
similarityMeasure = new PearsonSimilarity();
userSevenNeighboursAndSimilarities = userSeven.getNearestNeighboursAndSimilarities(3, allUsers.Values.ToList(), similarityMeasure);
Console.WriteLine("");
Console.WriteLine("Neighbours and similarities using PEARSON: ");
outputNeighbours(userSevenNeighboursAndSimilarities);
similarityMeasure = new CosineSimilarity();
userSevenNeighboursAndSimilarities = userSeven.getNearestNeighboursAndSimilarities(3, allUsers.Values.ToList(), similarityMeasure);
Console.WriteLine("");
Console.WriteLine("Neighbours and similarities using COSINE: ");
outputNeighbours(userSevenNeighboursAndSimilarities);
}
private static void singlePredictionDiagnostic(Dictionary <int, User> allUsers)
{
var userSeven = allUsers[7];
ISimilarity similarityMeasure = new PearsonSimilarity();
var userSevenNeighboursAndSimilarities = userSeven.getNearestNeighboursAndSimilarities(3, allUsers.Values.ToList(), similarityMeasure);
Console.WriteLine("");
IRatingPrediction weightedAvgPrediction = new WeightedAveragePrediction();
Console.WriteLine("User no. 7 predicted rating of product no. 101 using pearson is "
+ weightedAvgPrediction.predictRating(101, userSevenNeighboursAndSimilarities));
Console.WriteLine("User no. 7 predicted rating of product no. 103 using pearson is "
+ weightedAvgPrediction.predictRating(103, userSevenNeighboursAndSimilarities));
Console.WriteLine("User no. 7 predicted rating of product no. 106 using pearson is "
+ weightedAvgPrediction.predictRating(106, userSevenNeighboursAndSimilarities));
//INCORRECT, CHECK THIS ONE
var userFour = allUsers[4];
var userFourNeighboursAndSimilarities = userFour.getNearestNeighboursAndSimilarities(3, allUsers.Values.ToList(), similarityMeasure);
Console.WriteLine("");
Console.WriteLine("User no. 4 predicted rating of product no. 101 using pearson is "
+ weightedAvgPrediction.predictRating(101, userFourNeighboursAndSimilarities));
Console.WriteLine("");
Console.WriteLine("USER NO. 7 RATED ITEM 106 AS A 2.8: REDOING PREDICTIONS OF ITEMS 101 and 103");
userSeven.ratings[106] = 2.8;
userSevenNeighboursAndSimilarities = userSeven.getNearestNeighboursAndSimilarities(3, allUsers.Values.ToList(), similarityMeasure);
Console.WriteLine("User no. 7 predicted rating of product no. 101, with rating of product no. 106 set to 2.8 using pearson is "
+ weightedAvgPrediction.predictRating(101, userSevenNeighboursAndSimilarities));
Console.WriteLine("User no. 7 predicted rating of product no. 103, with rating of product no. 106 set to 2.8 using pearson is "
+ weightedAvgPrediction.predictRating(103, userSevenNeighboursAndSimilarities));
Console.WriteLine("");
Console.WriteLine("USER NO. 7 NOW RATED ITEM 106 AS A 5: REDOING PREDICTIONS OF ITEMS 101 and 103");
userSeven.ratings[106] = 5;
userSevenNeighboursAndSimilarities = userSeven.getNearestNeighboursAndSimilarities(3, allUsers.Values.ToList(), similarityMeasure);
Console.WriteLine("User no. 7 predicted rating of product no. 101, with rating of product no. 106 updated to 5 using pearson is "
+ weightedAvgPrediction.predictRating(101, userSevenNeighboursAndSimilarities));
Console.WriteLine("User no. 7 predicted rating of product no. 103, with rating of product no. 106 updated to 5 using pearson is "
+ weightedAvgPrediction.predictRating(103, userSevenNeighboursAndSimilarities));
}
private static void similarityMeasuresDiagnostic(Dictionary <int, User> allUsers)
{
ISimilarity similarityMeasurer = new PearsonSimilarity();
var testUser1 = allUsers[1];
var testUser3 = allUsers[3];
var overlappingProductsVectors = testUser1.getRatingsVectors(testUser3, similarityMeasurer.canHandleSparseData());
var testUser1Vector = overlappingProductsVectors[testUser1.userId];
var testUser3Vector = overlappingProductsVectors[testUser3.userId];
Console.WriteLine("User 1 ratings vector: " + testUser1Vector.ToString());
Console.WriteLine("User 3 ratings vector: " + testUser3Vector.ToString());
Console.WriteLine("Pearson similarity between user 1 and user 3: "
+ similarityMeasurer.computeSimilarity(testUser1Vector, testUser3Vector));
}
private static void nDimensionalRecommendationDiagnostic(Dictionary <int, User> sortedUserItems)
{
var user186 = sortedUserItems[186];
ISimilarity pearsonSimilarity = new PearsonSimilarity();
IRatingPrediction weightedAvgPrediction = new WeightedAveragePrediction();
int neighboursAmount = 25;
RecommendationCreator recommendationCreator = new RecommendationCreator(pearsonSimilarity, weightedAvgPrediction, neighboursAmount);
int recommendationAmount = 8;
int minimumAmountOfNeighbourProductOccurences = 0;
var top8Recommendations = recommendationCreator.getListOfTopPredictedRatings(recommendationAmount, minimumAmountOfNeighbourProductOccurences, sortedUserItems, user186);
Console.WriteLine("");
Console.WriteLine("Top 8 recommendations for user 186 with " + neighboursAmount + " neighbours and minimumAmountOfNeighbourProductOccurences = 0: ");
Console.WriteLine("-------------------------------------");
Console.WriteLine("");
int counter = 0;
top8Recommendations.ForEach(x => Console.WriteLine("Top " + (++counter) + " product: Movie no. " + x.Key + ", predicted rating " + x.Value));
Console.WriteLine("");
recommendationAmount = 8;
minimumAmountOfNeighbourProductOccurences = 3;
top8Recommendations = recommendationCreator.getListOfTopPredictedRatings(recommendationAmount, minimumAmountOfNeighbourProductOccurences, sortedUserItems, user186);
Console.WriteLine("");
Console.WriteLine("Top 8 recommendations for user 186 with " + neighboursAmount + " neighbours and minimumAmountOfNeighbourProductOccurences = 3: ");
Console.WriteLine("-------------------------------------");
Console.WriteLine("");
counter = 0;
top8Recommendations.ForEach(x => Console.WriteLine("Top " + (++counter) + " product: Movie no. " + x.Key + ", predicted rating " + x.Value));
}