// public Boolean DoNormalizeOnDomainLevel { get; set; } = true;
/// <summary>
/// Gets the preset tfidf.
/// </summary>
/// <param name="documentLimit">The document limit.</param>
/// <param name="log">The log.</param>
/// <returns></returns>
public static DocumentRankingMethod GetPreset_TFIDF(Int32 documentLimit, ILogBuilder log)
{
DocumentRankingMethod output = new DocumentRankingMethod();
ScoreModel scoreModel = new ScoreModel();
ScoreModelTermWeightFactor twf = new ScoreModelTermWeightFactor();
twf.TermWeightModel = new Weighting.FeatureWeightModel();
twf.TermWeightModel.LocalFunction = new TermFrequencyFunction();
FeatureWeightFactor featureWeightFactor = new FeatureWeightFactor();
featureWeightFactor.Settings = new GlobalFunctionSettings();
featureWeightFactor.Settings.functionName = nameof(IDFElement);
featureWeightFactor.Settings.weight = 1.0;
featureWeightFactor.Deploy(log);
twf.TermWeightModel.GlobalFactors.Add(featureWeightFactor);
twf.weight = 1.0;
scoreModel.Factors.Add(twf);
output.query = new DocumentSelectQuery();
output.query.SizeLimit = documentLimit;
output.model = scoreModel;
return(output);
}
/// <summary>Configures a global function based on Gravity moment</summary>
/// <remarks><para>What it will do?</para></remarks>
/// <param name="l">"Lambda factor of IGM"</param>
/// <param name="weight">Weigth associated with the function</param>
/// <param name="removeExisting">If any existing global factor should be removed</param>
/// <seealso cref="aceOperationSetExecutorBase"/>
public void aceOperation_setGlobalIGMWeight(
[Description("Lambda factor of IGM")] Double l = 7.0,
[Description("Weigth associated with the function")] Double weight = 1.0,
[Description("If any existing global factor should be removed")] Boolean removeExisting = false)
{
if (removeExisting)
{
data.GlobalFactors.Clear();
}
FeatureWeightFactor model = new FeatureWeightFactor();
model.Settings.functionName = nameof(IGMElement);
model.Settings.l = l;
model.Settings.weight = weight;
data.GlobalFactors.Add(model);
}
/// <summary>Configures a global function based on Term Discrimination Power</summary>
/// <remarks><para>It will add specified global factor, optionally if will remove any existing global factors</para></remarks>
/// <param name="factor">What factor should be added</param>
/// <param name="weight">Weigth associated with the function</param>
/// <param name="removeExisting">If any existing global factor should be removed</param>
/// <seealso cref="aceOperationSetExecutorBase"/>
public void aceOperation_setGlobalTDPWeight(
[Description("What factor should be added")] TDPFactor factor = TDPFactor.chi,
[Description("Weigth associated with the function")] Double weight = 1.0,
[Description("If any existing global factor should be removed")] Boolean removeExisting = false)
{
if (removeExisting)
{
data.GlobalFactors.Clear();
}
FeatureWeightFactor model = new FeatureWeightFactor();
model.Settings.functionName = nameof(CollectionTDPElement);
model.Settings.flags.Add(factor.ToString());
model.Settings.weight = weight;
data.GlobalFactors.Add(model);
}
/// <summary>
/// Configures a global function in the feature weighting model - supports: ICF, ICSd, IDF, IGM, mIDF
/// </summary>
/// <param name="function">Name of function elemenet</param>
/// <param name="weight">Weigth associated with the function</param>
/// <param name="IDF">How IDF should be computed</param>
/// <param name="removeExisting">if set to <c>true</c> [remove existing].</param>
/// <remarks>
/// It will add specified global factor, optionally if will remove any existing global factors
/// </remarks>
/// <seealso cref="aceOperationSetExecutorBase" />
public void aceOperation_setGlobalWeight(
[Description("Name of function elemenet")] String function = "IDFElement",
[Description("Weigth associated with the function")] Double weight = 1.0,
[Description("How IDF should be computed")] String flags = "logPlus",
[Description("If any existing global factor should be removed")] Boolean removeExisting = false)
{
if (removeExisting)
{
data.GlobalFactors.Clear();
}
FeatureWeightFactor model = new FeatureWeightFactor();
model.Settings.functionName = function;
model.Settings.flags.AddRange(flags.SplitSmart(","));
model.Settings.weight = weight;
data.GlobalFactors.Add(model);
}
/// <summary>
/// Gets default configuration
/// </summary>
/// <returns></returns>
public static PlanesMethodSettings GetDefaultSettings()
{
PlanesMethodSettings output = new PlanesMethodSettings();
output.entityMethod.instructions.Add(DocumentRenderInstruction.GetDescriptionInstruction());
output.entityMethod.instructions.Add(DocumentRenderInstruction.GetTitleInstruction());
output.entityMethod.instructions.Add(DocumentRenderInstruction.GetBodyTextInstruction());
// output.entityMethod.blenderOptions = DocumentBlenderFunctionOptions.binaryAggregation | DocumentBlenderFunctionOptions.pageLevel;
output.entityMethod.filterFunctionName = ""; // nameof(DocumentEntropyFunction);
output.entityMethod.filterLimit = 5;
output.corpusMethod.stemmer = nameof(EnglishStemmer);
output.corpusMethod.tokenizer = nameof(TokenizerBasic);
output.corpusMethod.transliterationRuleSetId = "";
#region PREPARE Weighting model
var weightModel = new FeatureWeightModel();
weightModel.LocalFunction = new Weighting.Local.TermFrequencyFunction();
var globalFactor = new FeatureWeightFactor();
globalFactor.Settings.functionName = nameof(IDFElement);
weightModel.GlobalFactors.Add(globalFactor);
output.corpusMethod.WeightModel = weightModel;
#endregion
var featureFilter = new FeatureFilter();
featureFilter.limit = 8000;
//featureFilter.
//featureFilter.functionSettings = new GlobalFunctionSettings();
//featureFilter.functionSettings.functionName = nameof(CollectionTDPElement);
//featureFilter.functionSettings.weight = 1.0;
//featureFilter.functionSettings.flags.Add(Weighting.Metrics.TDPFactor.chi.ToString());
output.corpusMethod.filter = featureFilter;
/*
* output.vectorMethod.constructor = new Feature.Settings.FeatureVectorConstructorSettings();
* dimensionSpecification dimSpec = new dimensionSpecification();
* dimSpec.functionName = nameof(CosineSimilarityFunction);
* dimSpec.type = FeatureVectorDimensionType.similarityFunction;
* output.vectorMethod.constructor.labelDimensions.Add(dimSpec);
*/
//output.vectorMethod.constructor = new Feature.Settings.FeatureVectorConstructorSettings();
//dimensionSpecification dimSpec = new dimensionSpecification();
////dimSpec.functionName = nameof(CosineSimilarityFunction);
//dimSpec.type = FeatureVectorDimensionType.directTermWeight;
//output.vectorMethod.constructor.featureDimensions.Add(dimSpec);
output.featureMethod.classifierSettings.type = Classifiers.ClassifierType.multiClassSVM;
output.featureMethod.classifierSettings.lossFunctionForTraining = Accord.MachineLearning.VectorMachines.Learning.Loss.L2;
/*
* output.featureMethod.classifierSettings.type = Classifiers.ClassifierType.kNearestNeighbors;
* output.featureMethod.classifierSettings.lossFunctionForTraining = Accord.MachineLearning.VectorMachines.Learning.Loss.L2;
* output.featureMethod.classifierSettings.kNN_k = 4;
*/
return(output);
}