Solr 3.3.0 - Increase field weight in search - boost field priority

solr field edismax
Mcgyver83 picture Mcgyver83 · Oct 12, 2011 · Viewed 9.6k times · Source

I'm indexing with solr 3.3.0 these docs:

<book id="bk101">
        <keywords>----</keywords>
        <title>----</title>
        <owner>----</owner>
        <artist>abcde</artist>
    </book>
    <book id="bk102">
        <keywords>----</keywords>
        <title>----</title>
        <owner>abcde</owner>
        <artist>----</artist>
    </book>
    <book id="bk103">
        <keywords>----</keywords>
        <title>----</title>
        <owner>----</owner>
        <artist>abc</artist>
    </book>
    <book id="bk104">
        <keywords>----</keywords>
        <title>----</title>
        <owner>abc</owner>
        <artist>----</artist>
    </book>
    <book id="bk105">
        <keywords>abcde</keywords>
        <title>----</title>
        <owner>----</owner>
        <artist>----</artist>
    </book>
    <book id="bk106">
        <keywords>abc</keywords>
        <title>----</title>
        <owner>----</owner>
        <artist>----</artist>
    </book>
    <book id="bk107">
        <keywords>----</keywords>
        <title>abcde</title>
        <owner>----</owner>
        <artist>----</artist>
    </book>
    <book id="bk108">
        <keywords>----</keywords>
        <title>abc</title>
        <owner>----</owner>
        <artist>----</artist>
    </book>

Schema.xml

<types>
    <fieldType name="text" class="solr.TextField" positionIncrementGap="100" omitNorms="false">
        <tokenizer class="solr.WhitespaceTokenizerFactory" />
        <filter class="solr.WordDelimiterFilterFactory" generateWordParts="1" generateNumberParts="1" catenateWords="1" catenateNumbers="1" catenateAll="0" splitOnCaseChange="0" splitOnNumerics="0" />
        <filter class="solr.LowerCaseFilterFactory" />
        <filter class="solr.StopFilterFactory" words="stopwords.txt" ignoreCase="true" />
        <filter class="solr.ISOLatin1AccentFilterFactory" />
        <filter class="solr.RemoveDuplicatesTokenFilterFactory" />
        <filter class="solr.EdgeNGramFilterFactory" minGramSize="1" maxGramSize="15" side="front" />
        <filter class="solr.EdgeNGramFilterFactory" minGramSize="1" maxGramSize="15" side="back" />
    </fieldType>

    <fieldType name="text_original" class="solr.TextField" positionIncrementGap="100" omitNorms="false">
        <tokenizer class="solr.KeywordTokenizerFactory" />
        <filter class="solr.LowerCaseFilterFactory" />
    </fieldType>

    <fieldType name="uuid" class="solr.UUIDField" indexed="true" />
    <fieldType name="string" class="solr.StrField" sortMissingLast="true" omitNorms="true" />
    <fieldType name="float" class="solr.FloatField" omitNorms="true" />
    <fieldType name="date" class="solr.DateField" />
</types>

<fields>
    <field name="id" type="uuid" indexed="true" stored="true" default="NEW" />
    <field name="book_id" type="string" indexed="true" stored="true" />
    <field name="keywords" type="text" indexed="true" stored="true" />
    <field name="keywords_ex" type="text_original" indexed="true" stored="true" />
    <field name="title" type="text" indexed="true" stored="true" />
    <field name="title_ex" type="text_original" indexed="true" stored="true" />
    <field name="owner" type="text" indexed="true" stored="true" />
    <field name="owner_ex" type="text_original" indexed="true" stored="true" />
    <field name="artist" type="text" indexed="true" stored="true" />
    <field name="artist_ex" type="text_original" indexed="true" stored="true" />

    <copyField source="title" dest="title_ex" />
    <copyField source="keywords" dest="keywords_ex" />
    <copyField source="owner" dest="owner_ex" />
    <copyField source="artist" dest="artist_ex" />

</fields>

As you can see if I search for 'abc' solr returns all docs (I'm using solr.EdgeNGramFilterFactory for indexing); my achievement is to have as result documents sorted using this logic:

  1. title_ex
  2. title
  3. keywords_ex
  4. keywords
  5. owner_ex
  6. artist_ex
  7. owner
  8. artist

I tryed "^" lucene sintax and dismax but the response list isn't the one I supposed.

`http://localhost:8080/solr33a/core0/select?q=abc&defType=edismax&qf=title_ex^10%20title^8.0%20keywords_ex^6%20keywords^5.5%20owner_ex^1.2%20artist_ex^0.8%20owner^0.5%20artist^0.2&fl=*,score

Result:

<doc>
            <float name="score">2.3862944</float>
            <str name="artist">----</str>
            <str name="artist_ex">----</str>
            <str name="book_id">bk108</str>
            <str name="id">2cc5d478-6901-4777-abc9-680fd959ef90</str>
            <str name="keywords">----</str>
            <str name="keywords_ex">----</str>
            <str name="owner">----</str>
            <str name="owner_ex">----</str>
            <str name="title">abc</str>
            <str name="title_ex">abc</str>
        </doc>
        <doc>
            <float name="score">1.4317766</float>
            <str name="artist">----</str>
            <str name="artist_ex">----</str>
            <str name="book_id">bk106</str>
            <str name="id">e12683a2-faff-4d86-8107-7406491f4f89</str>
            <str name="keywords">abc</str>
            <str name="keywords_ex">abc</str>
            <str name="owner">----</str>
            <str name="owner_ex">----</str>
            <str name="title">----</str>
            <str name="title_ex">----</str>
        </doc>
        <doc>
            <float name="score">0.3288517</float>
            <str name="artist">----</str>
            <str name="artist_ex">----</str>
            <str name="book_id">bk107</str>
            <str name="id">a6a4a014-ce94-4257-a215-c1a64aa41cf5</str>
            <str name="keywords">----</str>
            <str name="keywords_ex">----</str>
            <str name="owner">----</str>
            <str name="owner_ex">----</str>
            <str name="title">abcde</str>
            <str name="title_ex">abcde</str>
        </doc>
        <doc>
            <float name="score">0.28635535</float>
            <str name="artist">----</str>
            <str name="artist_ex">----</str>
            <str name="book_id">bk104</str>
            <str name="id">9b909c65-e56a-4407-a789-53a570a7ae40</str>
            <str name="keywords">----</str>
            <str name="keywords_ex">----</str>
            <str name="owner">abc</str>
            <str name="owner_ex">abc</str>
            <str name="title">----</str>
            <str name="title_ex">----</str>
        </doc>
        <doc>
            <float name="score">0.22608554</float>
            <str name="artist">----</str>
            <str name="artist_ex">----</str>
            <str name="book_id">bk105</str>
            <str name="id">dead87cc-f93b-4562-af32-4d9fb2613c7f</str>
            <str name="keywords">abcde</str>
            <str name="keywords_ex">abcde</str>
            <str name="owner">----</str>
            <str name="owner_ex">----</str>
            <str name="title">----</str>
            <str name="title_ex">----</str>
        </doc>
        <doc>
            <float name="score">0.19090356</float>
            <str name="artist">abc</str>
            <str name="artist_ex">abc</str>
            <str name="book_id">bk103</str>
            <str name="id">306a252c-a0b5-474d-b55d-a25740d063b4</str>
            <str name="keywords">----</str>
            <str name="keywords_ex">----</str>
            <str name="owner">----</str>
            <str name="owner_ex">----</str>
            <str name="title">----</str>
            <str name="title_ex">----</str>
        </doc>
        <doc>
            <float name="score">0.020553231</float>
            <str name="artist">----</str>
            <str name="artist_ex">----</str>
            <str name="book_id">bk102</str>
            <str name="id">a684de0c-b286-4d9e-bd68-d5305afeee76</str>
            <str name="keywords">----</str>
            <str name="keywords_ex">----</str>
            <str name="owner">abcde</str>
            <str name="owner_ex">abcde</str>
            <str name="title">----</str>
            <str name="title_ex">----</str>
        </doc>
        <doc>
            <float name="score">0.008221293</float>
            <str name="artist">abcde</str>
            <str name="artist_ex">abcde</str>
            <str name="book_id">bk101</str>
            <str name="id">30a0f9de-1224-49d2-90aa-41f57af4956c</str>
            <str name="keywords">----</str>
            <str name="keywords_ex">----</str>
            <str name="owner">----</str>
            <str name="owner_ex">----</str>
            <str name="title">----</str>
            <str name="title_ex">----</str>
        </doc>

Answer

Jayendra picture Jayendra · Oct 12, 2011

As you want to search across multiple fields with variable weightage, configure your request handler to use edismax query parser.
dismax allows you to search across fields and add weights to each field.

Example title match has weight 1, while matches on author 0.8, so that document having title matches would appear at the top.

<requestHandler name="search" class="solr.SearchHandler" default="true">
 <lst name="defaults">
   <str name="echoParams">explicit</str>
   <str name="defType">edismax</str>
   <str name="qf">
      title^1 author^0.8
   </str>
   <str name="q.alt">*:*</str>
   <str name="rows">10</str>
   <str name="fl">*,score</str>
 </lst>
</requestHandler>

Related questions

What is omitNorms and version field in solr schema?

I am not understanding when to use omitNorms="true". I read 2-3 links but still I am not clear with its meaning. what does it mean "Set to true to omit the norms associated with this field (this disables length …

Read More
Solr search query for dynamic fields indexed

I am implementing the solr search in my project . I have one question regarding how do i search a dynamic fields that is created in a solr index Eg:- this is the tag that is formed in the index …

Read More
Solr vs. ElasticSearch

What are the core architectural differences between these technologies? Also, what use cases are generally more appropriate for each?

Read More