In faceted search (also known as out-of-the-box search), information can be accessed based upon dynamic classification. Faceted search uses a faceted classification system in which indexed information can be accessed and explored from various dimensions using filters. The facets allow you to filter data through multiple paths and different ordering, and also click on any combination of facets to get the desired results.

Features and Facets

SearchBlox can perform a customized search using this faceted search functionality. The default facets available are collections, keywords, file type and date. You can filter the search results by selecting the required combination of facets.

How to Perform a Faceted Search

The SearchBlox faceted search can be accessed through the following URL:
localhost:8080/searchblox/plugin/index.html

Provide the search term in the search field and press enter. The search results page with the facets will appear in the left-hand menu as shown in the following screenshot:

Facet Filters

The default facets for filters such as file type, keyword and modified date can be found in the left-hand menu. After selecting each filter, the facets will appear selected with x mark to deselect the same. By clicking the x next to the facet, it can be removed from the search results as shown in the following image.

• File Type lists the file types available in the search results for filtering.
• Keywords lists the words that are the keywords of the search, other than the search term results.
• Modified Date has five options ranging from week to year, along with customized date range.
  You can also add collection as a facet by making changes in ../plugin/data/facet.json
• Collection lists the collections that have the search results for the given search term.
  Similarly, you can also add other meta tags as facets.

Adding a Facet in SearchBlox

SearchBlox also supports Collection, Title, Description, Language, Author and Subject as facets. The last two facets (author and subject) are applicable for only PDF files.
facet.js is the configuration file that provides options that can be configured to provide search results as required. The path for the facet.js file, after placing the plugin folder in SearchBlox, is ../webapps/searchblox/plugin/facet.js. To add a new facet filter, define the field name, display name and the size of the filter values to be shown in facet.js. Facet filters can be string, number or date types. Except for the predefined facets such as keywords, title, description, size or date, if you need to add more custom facets, you can add them in facet.js (example: facet field: "group").

Fields in facets.js

The contents of facet.js are shown here.

window.facets = {
    "facets": [{
            "field": "colname",
            "display": "Collection Name",
            "size": "10"
        },
        {
            "field": "contenttype",
            "display": "File Type",
            "size": "10"
        },
        {
            "field": "keywords",
            "display": "keywords",
            "size": "10"
        },
        {
            "field": "lastmodified",
            "display": "Last Modified",
            "dateRange": [{
                    "name": "Last 24 hours",
                    "calendar": "days",
                    "value": "1"
                },
                {
                    "name": "Past Week",
                    "calendar": "days",
                    "value": "7"
                },
                {
                    "name": "Past Month",
                    "calendar": "months",
                    "value": "1"
                },
                {
                    "name": "Past Year",
                    "calendar": "years",
                    "value": "1"
                }
            ]
        }
    ],
    "collection": [],
    "sortBtns": [{
            "field": "lastmodified",
            "display": "Sort by date"
        },
        {
            "field": "relevance",
            "display": "Sort by relevance"
        }
    ],
    "facetFiltersOrder": [
      "colname","keywords"
    ],
    "facetFiltersType": "AND",
    "sortDir": "desc",
    "matchAny": "off",
    "pageSize": "10",
    "showAutoSuggest": "true",
    "defaultCname": "",
    "adsDisplay": "true",
    "featuredResultsCount": "3",
    "predictSearch": "false",
    "predictResultSize": "3",
    "customDateDisplayText":"Custom",
    "customDateEnable":"true",
    "relatedQuery": true,
    "relatedQueryFields": {
        "apikey": "BFC82BE93A351C018CB27C32E1138047",
        "field": "content",
        "operator": "and",
        "limit": "5",
        "terms": "10",
        "type": "phrase",
        "col": ""
    },
    "dataToBeDisplayed": {
        "1": {
            "title": "Title",
            "description": "Description"
        },
        "other": {
            "description": "Description"
        },
        "displayAll": true
    },
    "tune": {
        "enable": "true",
        "tune.0": "5",
        "tune.1": "3",
        "tune.2": "2",
        "tune.3": "5",
        "tune.4": "180",
        "tune.5": "10"
    },
    "defaultType": "",
    "pluginDomain": ""
};

• If you need to add a meta field as facet, then the field has to be mapped in mapping.json. For CSV, database, MongoDB and AWS collections, the values need to be mapped in csv.json, jdbc.json, mongodb.json, amazonS3.json respectively here: ../webapps/searchblox/WEB-INF/.
• If you need to add a facet field for type number or date, the field needs to be updated in ../webapps/searchblox/WEB-INF/mapping.json for web and file collections. For CSV, database, MongoDB and AWS collections, the values need to be mapped in csv.json, jdbc.json, mongodb.json, amazonS3.json respectively here: ../webapps/searchblox/WEB-INF/.
  ref: Custom Fields in Search

Sorting Search Results

In faceted search, users can sort the search results based on date and relevance using the drop-down provided at the top of the search results and for the search box. By default, it would be in descending order.
The ascending or descending order of sorting can also be given using parameters in the URL. The user can also specify a custom field for sorting using facet.js.

AutoSuggest

Autosuggest/Autocomplete is enabled by default in faceted search for past queries. This feature can be disabled from the search settings if required.
AutoSuggest

Passing Parameters Through Faceted Search URL

In faceted search, filter search results by passing the parameters along with the query as shown in the following URL:
http://localhost:8080/searchblox/plugin/index.html?query=news AND keywords:cnn
More than one parameter can also be passed through the URL:
http://localhost:8080/searchblox/plugin/index.html?query=news AND keywords:cnn AND content:world

Also, it is possible to pass filter, collection, pagination, sort, search type, tune and many other parameters in faceted search. Please refer the following URL:
Example: http://localhost:8080/searchblox/plugin/index.html?page=1&pagesize=10&query=test&sort=lastmodified&sortdir=desc&adsDisplay=true&f.colname.size=10&f.contenttype.size=10&f.keywords.size=10&f.lastmodified.range=[2019-06-20T16:00:48TO*]&f.lastmodified.range=[2019-06-14T16:00:48TO*]&f.lastmodified.range=[2019-05-21T16:00:48TO*]&f.lastmodified.range=[2018-06-21T16:00:48TO*]&facet.field=colname&facet.field=contenttype&facet.field=keywords&facet.field=lastmodified&predictResultSize=3&predictSearch=false&relatedQuery=true&tune=true&tune.0=5&tune.1=3&tune.2=2&tune.3=5&tune.4=180&tune.5=10

Related Queries

In our latest plugin in SearchBlox 9.0 it is possible to get related queries for the search term provided. Please enable the settings in ../plugin/facet.js and provide your API key from SearchBlox Admin console -> license page as shown here

"relatedQuery": true,
    "relatedQueryFields": {
        "apikey": "BFC82BE93A351C018CB27C32E1138047",
        "field": "content",
        "operator": "and",
        "limit": "5",
        "terms": "10",
        "type": "phrase",
        "col": ""
    },

Providing relatedQuery as "false" will disable this feature.

Predicted Results

SearchBlox 9.0 has a new feature predicted results in the faceted search page. It would bring about predicated results for the particular search term provided
To enable this feature please provide predictSearch field value as true as shown. You can also specify the result size in ../plugin/facet.js as shown here.

"predictSearch": "true",
    "predictResultSize": "3",