/// <summary>
/// Trains a model using SAR.
/// </summary>
/// <param name="settings">The training settings</param>
/// <param name="usageEvents">The usage events to use for training</param>
/// <param name="catalogItems">The catalog items to use for training</param>
/// <param name="uniqueUsersCount">The number of users in the user id index file.</param>
/// <param name="uniqueUsageItemsCount">The number of usage items in the item id index file</param>
/// <param name="cancellationToken">A cancellation token</param>
public IPredictorModel Train(ITrainingSettings settings,
IList <SarUsageEvent> usageEvents,
IList <SarCatalogItem> catalogItems,
int uniqueUsersCount,
int uniqueUsageItemsCount,
CancellationToken cancellationToken)
{
if (settings == null)
{
throw new ArgumentNullException(nameof(settings));
}
if (usageEvents == null)
{
throw new ArgumentNullException(nameof(usageEvents));
}
if (settings.EnableColdItemPlacement && catalogItems == null)
{
throw new ArgumentNullException(nameof(catalogItems));
}
if (uniqueUsersCount < 0)
{
var exception = new ArgumentException($"{nameof(uniqueUsersCount)} must be a positive integer");
_tracer.TraceWarning(exception.ToString());
throw exception;
}
if (uniqueUsageItemsCount < 0)
{
var exception = new ArgumentException($"{nameof(uniqueUsageItemsCount)} must be a positive integer");
_tracer.TraceWarning(exception.ToString());
throw exception;
}
cancellationToken.ThrowIfCancellationRequested();
using (TlcEnvironment environment = new TlcEnvironment(verbose: true))
{
_detectedFeatureWeights = null;
try
{
environment.AddListener <ChannelMessage>(ChannelMessageListener);
IHost environmentHost = environment.Register("SarHost");
// bind the cancellation token to SAR cancellation
using (cancellationToken.Register(() => { environmentHost.StopExecution(); }))
{
_tracer.TraceInformation("Starting training model using SAR");
return(TrainModel(environmentHost, settings, usageEvents, catalogItems, uniqueUsersCount,
uniqueUsageItemsCount));
}
}
finally
{
environment.RemoveListener <ChannelMessage>(ChannelMessageListener);
}
}
}
/// <summary>
/// Trains a model using SAR.
/// </summary>
/// <param name="settings">The training settings</param>
/// <param name="usageEvents">The usage events to use for training</param>
/// <param name="catalogItems">The catalog items to use for training</param>
/// <param name="uniqueUsersCount">The number of users in the user id index file.</param>
/// <param name="uniqueUsageItemsCount">The number of usage items in the item id index file</param>
public IPredictorModel Train(ITrainingSettings settings,
IList <SarUsageEvent> usageEvents,
IList <SarCatalogItem> catalogItems,
int uniqueUsersCount,
int uniqueUsageItemsCount)
{
return(Train(settings, usageEvents, catalogItems, uniqueUsersCount, uniqueUsageItemsCount,
CancellationToken.None));
}
/// <summary>
/// Trains a model using SAR.
/// </summary>
/// <param name="settings">The training settings</param>
/// <param name="usageEvents">The usage events to use for training</param>
/// <param name="catalogItems">The catalog items to use for training</param>
/// <param name="featureNames">The names of the catalog items features, in the same order as the feature values in the catalog</param>
/// <param name="uniqueUsersCount">The number of users in the user id index file.</param>
/// <param name="uniqueUsageItemsCount">The number of usage items in the item id index file</param>
/// <param name="catalogFeatureWeights">The computed catalog items features weights (if relevant)</param>
public IPredictorModel Train(ITrainingSettings settings,
IList <SarUsageEvent> usageEvents,
IList <SarCatalogItem> catalogItems,
string[] featureNames,
int uniqueUsersCount,
int uniqueUsageItemsCount,
out IDictionary <string, double> catalogFeatureWeights)
{
return(Train(settings, usageEvents, catalogItems, featureNames, uniqueUsersCount, uniqueUsageItemsCount,
out catalogFeatureWeights, CancellationToken.None));
}
/// <summary>
/// Trains a model using SAR.
/// </summary>
private IPredictorModel TrainModel(
IHost environment,
ITrainingSettings settings,
IList <SarUsageEvent> usageItems,
IList <SarCatalogItem> catalogItems,
int uniqueUsersCount,
int uniqueUsageItemsCount)
{
IDataView catalog = null;
if (settings.EnableColdItemPlacement)
{
SarCatalogItem item = catalogItems.FirstOrDefault();
int featuresVectorSize = item?.FeatureVector.Length ?? 0;
_tracer.TraceInformation($"Found catalog item features vector size of {featuresVectorSize}");
// check if the catalog has items with any features. if not, there is no point to compute 'cold' items recommendations
if (featuresVectorSize > 0)
{
_tracer.TraceInformation($"Creating catalog item schema using the features vector size of {featuresVectorSize}");
var catalogItemSchema = SchemaDefinition.Create(typeof(SarCatalogItem));
catalogItemSchema["Item"].ColumnType = new KeyType(DataKind.U4, 1, uniqueUsageItemsCount);
catalogItemSchema["Features"].ColumnType = new VectorType(TextType.Instance, featuresVectorSize);
catalog = environment.CreateDataView(catalogItems, catalogItemSchema);
}
}
_tracer.TraceInformation("Creating usage item schema");
var usageItemSchema = SchemaDefinition.Create(typeof(SarUsageEvent));
usageItemSchema["user"].ColumnType = new KeyType(DataKind.U4, 1, uniqueUsersCount);
usageItemSchema["Item"].ColumnType = new KeyType(DataKind.U4, 1, uniqueUsageItemsCount);
// create a usage data view
IDataView usage = environment.CreateDataView(usageItems, usageItemSchema);
// set the similarity function factory
ISimCalculatorFactory simCalculatorFactory;
switch (settings.SimilarityFunction)
{
case SimilarityFunction.Jaccard:
simCalculatorFactory = new JaccardSimilarityCalculator.Arguments
{
threshold = settings.SupportThreshold
};
break;
case SimilarityFunction.Lift:
simCalculatorFactory = new LiftSimilarityCalculator.Arguments
{
threshold = settings.SupportThreshold
};
break;
case SimilarityFunction.Cooccurrence:
simCalculatorFactory = new CoOccurrenceSimilarityCalculator.Arguments
{
threshold = settings.SupportThreshold
};
break;
default:
var exception = new ArgumentException($"Unknown similarity function '{settings.SimilarityFunction}'");
_tracer.TraceError(exception.ToString());
throw exception;
}
// prepare SAR trainer's input arguments
var trainInput = new SAR.Input
{
TrainingData = usage,
CatalogData = catalog,
Calculator = simCalculatorFactory,
Backfill = settings.EnableBackfilling,
ItemColumn = "Item",
UserColumn = "user",
ColdToCold = settings.EnableColdToColdRecommendations,
MaxColdItems = 10,
MultiValueFeatures = true
};
_tracer.TraceInformation("Training a SAR predictor using the usage and catalog data");
SAR.Output trainOutput = SAR.Train(environment, trainInput);
return(trainOutput.PredictorModel);
}
/// <summary>
/// Trains a model using SAR.
/// </summary>
/// <param name="settings">The training settings</param>
/// <param name="usageEvents">The usage events to use for training</param>
/// <param name="catalogItems">The catalog items to use for training</param>
/// <param name="featureNames">The names of the catalog items features, in the same order as the feature values in the catalog</param>
/// <param name="uniqueUsersCount">The number of users in the user id index file.</param>
/// <param name="uniqueUsageItemsCount">The number of usage items in the item id index file</param>
/// <param name="catalogFeatureWeights">The computed catalog items features weights (if relevant)</param>
/// <param name="cancellationToken">A cancellation token</param>
public IPredictorModel Train(ITrainingSettings settings,
IList <SarUsageEvent> usageEvents,
IList <SarCatalogItem> catalogItems,
string[] featureNames,
int uniqueUsersCount,
int uniqueUsageItemsCount,
out IDictionary <string, double> catalogFeatureWeights,
CancellationToken cancellationToken)
{
if (settings == null)
{
throw new ArgumentNullException(nameof(settings));
}
if (usageEvents == null)
{
throw new ArgumentNullException(nameof(usageEvents));
}
if (settings.EnableColdItemPlacement && catalogItems == null)
{
throw new ArgumentNullException(nameof(catalogItems));
}
if (uniqueUsersCount < 0)
{
var exception = new ArgumentException($"{nameof(uniqueUsersCount)} must be a positive integer");
_tracer.TraceWarning(exception.ToString());
throw exception;
}
if (uniqueUsageItemsCount < 0)
{
var exception = new ArgumentException($"{nameof(uniqueUsageItemsCount)} must be a positive integer");
_tracer.TraceWarning(exception.ToString());
throw exception;
}
cancellationToken.ThrowIfCancellationRequested();
using (TlcEnvironment environment = new TlcEnvironment(verbose: true))
{
_detectedFeatureWeights = null;
try
{
environment.AddListener <ChannelMessage>(ChannelMessageListener);
IHost environmentHost = environment.Register("SarHost");
// bind the cancellation token to SAR cancellation
using (cancellationToken.Register(() => { environmentHost.StopExecution(); }))
{
_tracer.TraceInformation("Starting training model using SAR");
IPredictorModel model = TrainModel(environmentHost, settings, usageEvents, catalogItems, uniqueUsersCount,
uniqueUsageItemsCount);
catalogFeatureWeights = new Dictionary <string, double>();
if (_detectedFeatureWeights != null && featureNames != null)
{
if (_detectedFeatureWeights.Length == featureNames.Length)
{
for (int i = 0; i < featureNames.Length; i++)
{
catalogFeatureWeights[featureNames[i]] = _detectedFeatureWeights[i];
}
}
else
{
_tracer.TraceWarning(
$"Found a mismatch between number of feature names ({featureNames.Length}) and the number of feature weights ({_detectedFeatureWeights.Length})");
}
}
return(model);
}
}
finally
{
environment.RemoveListener <ChannelMessage>(ChannelMessageListener);
}
}
}
