More like this

MoreLikeThisQuery finds documents similar to a source document by extracting representative terms from its term vectors and turning them into a weighted boolean query.

var mlt = new MoreLikeThisQuery(
    docId: 42,
    fields: new[] { "title", "body" },
    parameters: new MoreLikeThisParameters
    {
        MinTermFreq    = 2,
        MinDocFreq     = 5,
        MaxQueryTerms  = 25,
        MinWordLength  = 3,
        BoostByScore   = true
    });

var hits = searcher.Search(mlt, topN: 10);

Parameters

Parameter	Default	Effect
`MinTermFreq`	`1`	Minimum occurrences in the source doc
`MinDocFreq`	`1`	Minimum docs containing the term
`MaxDocFreq`	`int.MaxValue`	Cap on doc frequency for common terms
`MaxQueryTerms`	`25`	Maximum terms in the generated query
`MinWordLength`	`3`	Minimum term length
`BoostByScore`	`true`	Weight terms by their TF-IDF score

Raise MinDocFreq to filter rare terms that dominate similarity. Lower MaxDocFreq to filter ultra-common terms.

More like this

Parameters

See also