// 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>
/// Defines dataset filter, based on current FeatureWeightFactor model
/// </summary>
/// <param name="remove">If true it will remove any existing ScoreModel Factors</param>
/// <param name="weight">Weight associated with the term weight based document score factor</param>
/// <remarks>
/// It will copy current FeatureWeight model as a ScoreModel factor for the DataSet filtration function
/// </remarks>
/// <seealso cref="aceOperationSetExecutorBase" />
public void aceOperation_setDocumentSelectionAddTWFactor(
[Description("If true it will remove any existing ScoreModel Factors")] Boolean remove = false,
[Description("Model definition name")] String modelName = "*",
[Description("Weight associated with the term weight based document score factor")] Double weight = 1.0,
[Description("Normalize score value on range from 0.0 to 1.0, across the sample")] Boolean normalize = true
)
{
if (remove)
{
data.ranking.model.Factors.Clear();
data.ranking.model.SerializedFactors.Clear();
}
ScoreModelTermWeightFactor twf = new ScoreModelTermWeightFactor();
twf.modelDefinitionFile = modelName;
// twf.TermWeightModel = weighting.data.weightModel; // toolkitSettings.corpusMethod.weightModel;
twf.weight = weight;
twf.doNormalize = normalize;
data.ranking.model.Factors.Add(twf);
}