Fields

Owned by Former user (Deleted)
Last updated: Jul 16, 2023 by Joyce Chu

In this article you will find:


Fields

The Fields section allows you to monitor which fields are being indexed from the data that is provided. You can view and edit how the data in these field is being used:  for faceting, indexing, stemming, or for keywords. Additionally, you can quickly see if the data in that field is being used as the Primary Key and/or a queryable field.


Navigating the Field Administration Page

1) To navigate to the Fields Administration Page, click on the "Workbench" tab in the top navigation.

2) Then, select "Data Configuration" from the left navigation and choose "Fields" from the subcategories.

3) The first search field allows you to search for the name of an existing Field. If you know the name as it appears in the data feed, or a keyword from the name, type it into the "Name Field."

4) You can also conduct a search based on the label of the Field. This name may differ from the corresponding field in the data field. 

5) Search based on the type of Field by selecting an option in the drop down.

6) To search for a Field by any tags associated with it, enter the tags here.

7) Once you have added all search criteria, click the "Search" button to see your results.

8) If you would like to edit one of the fields, you can do so by clicking the "pencil" icon. To delete a Field, click the "trash can" icon. To copy a Field, click the "paper" icon. 

9) To add a new Field, click on the "Add New Field " link.


Adding a New Field

1) First, chose a label for the Field you are creating by typing it into the "Label" field. The label may differ from the way the field appears in the data feed. 

2) Next, type the name of the field as it appears in the data being sent to Hawksearch. (Note: If you're sending fields with capital letters, these will need to be converted to lowercase for the Field Name)

3) Decide whether this field will be the Primary Key for the data set. The primary key should hold a one to one relationship with each item in the data, meaning each item has exactly one value corresponding to the field in the data. There should only be ONE primary key established for each data set being indexed. 

4) Select whether the information in the field is a Text Value (alphanumeric), Numeric only, or Boolean Value (only two possible values ex. Is_on_sale yes or no).

5) Choose the field type that will indicate how values in this field should be indexed using the drop down menu.

  • "Field values indexed "as is" AND are stemmed"- the value is indexed exactly as is and is also available for stemming. For example, brand name should be saved "as is" for use in facets, but may also be available in stemmed forms in queries.
  • "Field values are NOT stemmed"- These values will only be indexed as is. For example, SKU should only yield the exact value in search results. Or sizes should be saved exactly "as" is for use in facets.
  • "Field values are ONLY stemmed (search only)"- Stemming a value makes it available for searches of broader terms. For example, a product with the word "run" in the name field would also return similar items with "running" in the name. The values will not be used in facets so there is no need to index "as is".
  • "Stored only, not used for search nor facets"- This information is saved to the index, but is not intended for use in searches nor facets. For example, an image url that may be rendered in results, but will not be a basis for search.

6) Enter any tags that you would like to associate with the field. An example of applying tags to fields is to identify the source from where the field's values are fed. Another example is to label the field with the types of products using that field. However you decide to utilize the tags on the field, you will be able to see them in the list of fields and use them in the filters.


Advanced Options


There are a number of options for the configuration of a field that will become available based on the Field Type that is selected. Listed below are the number of options.


Query this field?

1) "Query this field?" determines whether the field will be available for searches. Hawksearch recommends turning on the minimum number of fields for query as possible that will yield desired results while minimizing redundancies. For example, if brand is included as part of the name field as a standard and name is set to query, there is no need to include brand in the list of fields that will query. To turn the query on, click the toggle to "on".

2) If needed, cascading queries can be done in multiple iterations. For example, if the "Title" field is the only searchable field and only 2 results are returned with the first search, a second query can be done that adds more fields for consideration. There can be up to 3 iterations of searching. Each field that is flagged as searchable with the field "Query this field?" will also have a value indicating which search iteration(s) the field should be considered for.

3) When you switch the "Query this Field?" Toggle to "on" you will be given a couple of additional options. The option first will prompt you to select an analyzer that will alter the indexing configuration from the drop down. If no analyzer is selected, the system will apply the default configuration to the field which is the Snowball Analyzer.

Types of Analyzers

  • Hawk Analyzer - Hawk's default analyzer; works like a Snowball Analyzer but also processes Synonyms (single and multi-word).
  • Standard Analyzer-  A JFlex-based grammar underlies this analyzer, recognizing the following lexical types: alphanumeric, acronyms, company names, email addresses, computer host-names, numbers and words with an interior apostrophe, serial numbers, IP addresses, Chinese and Japanese characters. StandardAnalyzer also includes stop-word removal.
  • Simple Analyzer- Divides text at non-letter characters and and applies lowercase to the text
  • Snowballer Analyzer- an aggressive stemming analyzer that will dilute down the content data to match based on broader terms
  • Stop Analyzer- Tolkenizes the data by removing special characters and adding white space, formats the data field to lower case and removes stop words from the data.
  • White Space Analyzer- Divides the text at white-space

4) The boost option allows you to rank the relevancy of fields available for queries, by assigning a value between 1 and 200. Ex. We could assign a boost of 50 to the name field and a boost of 1 to keywords. If an item matches in the "name" field and another item matches in the "keywords" field for a give search, the item that matched name will receive a higher score than the item that matched keywords because of the corresponding boost values. For more see Using Boost Values in Fields.

5) Partial Query Type indicates the way items will match when a partial search term is entered. When conducting a partial match, the value will be read from left to right. When "Wild Card" is selected, the searched value may start anywhere within the actual value to provide a match. When "Prefix" is selected, the searched value must start from the beginning of the actual value, but can cut off at any point.


More Advanced Options

1) Omit norms- Norms allow for field length normalization to help influence relevance. For example, if a document has 50,000 lines of text and another has 50 lines of text, the smaller document would score higher. Selecting "Omit Norms: on" will disable this option for this field. Enable this feature for fields that are very short (e.g. ids, names) and/or if there are many fields in the engine. 

2) Add phrases to "Did you Mean?" indicates that the data in this field should be used to find matches when the "Did you Mean?" feature is enabled and triggered. Selecting "on" indicated that phrases in the field should be included in the "Did you Mean?" matches.

3) Turning "Include in Dictionary" to "on" indicates that data in these fields should be included in the dictionary for the foundation of auto-correct.

4) When "Enable Sorting" is switched to "yes" the field is added as an option for the customer to sort the displayed results on landing pages only.

5) The "Best Fragment" feature creates a synopsis of a larger field that can be used for display. The types of field that utilize this feature are: Long Description, Title, Content (body content), or Name.  

6)The "Hide in Query Builder" option allows you to hide options for fields when building conditions and triggers using the workbench. When turned "on" the option to choose this field in the drop downs will not display.

7) In cases where there is a comparing products feature in place, turning the toggle to "on" would indicate the content in this field should be used for display.

8) "Include in Results" indicates that the data field will be used for product display on the results page.

9) "Skip from Custom" only applies when you choose to include the field in results. When customized columns in the data contain duplicate information from the standard columns, this repeated information will be removed from the API call.

10) The Strip HTML feature indicates to the engine that this field possess HTML and it should be removed. Some data fields are created with HTML to assist with the layout of the page. If the data field has HTML and it should be queried, the strip HTML feature should be enabled. If it is important that the HTML be returned with the content for display, the field should be added twice.

11) The option "Copy To" can be used, when you want to copy values associated with this field into another field that is registered in the fields section. This allows you to use the existing field with its current configuration but also makes the values in this field available in another field as well so it can be configured separately and differently for relevancy purposes. For example you can have "Brand" field set to copy to "Short Description" field. In this case you can have settings for brand field to only be "use for facets", however you can have the "short description" field marked as "use as stemmed" and the brand value now being available in that field will allow for brand keywords to be stemmed as well.


Recommendations

This section allows you to configure settings when the Recommendation feature is implemented.

1) "Field Mapping" specifies the way the field will be included in the display of the Recommendation feature.

Allowed types of Field Mappings

  • Brand
  • IsOnSale
  • BestFragment
  • ImageOverAltTag
  • ImageOverUrl
  • MaxPrice
  • MaxSalePrice
  • MinPrice
  • MinSalePrice
  • PostDate
  • Rating
  • Sku

2) The "Default Value" applies when a Recommendation strategy utilizes custom filters in the corresponding field. When there are zero or not enough items to display that will match that field. Items will then be matched to the default value specified here.


Saving a Field


Once you have made all of your selections, click the save button.


You should see your new Field displayed on the Facet Administration Page. The green check marks to the right of the fields indicate the Primary Key, when the field is used in the Query, and when the field is included in results.


Adding a Facet from the Field Administration Page

When viewing the list of fields in the Data Configuration Section of the workbench, you can also easily create a facet from a field by using the "new facet" button. Clicking the button will take you into the screen for creating a new facet with the field preselected. Fields that are already used in a facet will have a new button that links to the edit screen for the facet.