Search Terms

The Search Terms page in the Analysis section contains a search term family reporting feature. Use this feature to view statistics for search term hits in documents, and to view how each search term contributes to the overall population of hits for the family, along with other metrics.

Note: The counts display only base documents, and not renditions.

An example of the Search Terms page is shown in the following figure.

Example of the Search Terms page

You can download a search terms report to a spreadsheet (.csv file).

Administrators can make this feature available to group leaders and group members.

To work with search terms:

On the Case Home page, under Analysis, click Search Terms.

On the Search term family menu, select a search term family.

Optionally, refine the report by selecting any of the following:

Document set: By default, the report includes information for all documents in the case, unless you select a specific document population, such as a binder, issue, or sample.

Selecting a document set displays a new field where you can select an object for the document set.

Coding field: Measure the precision and recall of a search term family by selecting a coding field to view the effectiveness of your search terms. If you select a coding field, you must select one or more values on each of the following menus:

Tip: To select more than one value, press Shift while selecting values on the menu. To remove values, press Shift while selecting highlighted values on the menu.

Positive: A value identified as a positive mark for computing recall and precision. Terms with higher recall return a higher proportion of the documents coded positive in the selected document set. Terms with higher precision have a higher proportion of documents coded positive in the returned document set.

Negative: A value identified as a negative mark for computing recall and precision.

Click Apply.

Note: The counts display only base documents, and not renditions. 

Links appear in some columns on the Search Terms page. Click a link to open those documents on the Documents page.

To download the information on the Search Terms page to a spreadsheet (.csv file), click Download report.

The information described in the following table appears in the columns and rows on the Search Terms page.

Row or Column

Description

Term label

The label for the search term.

Term query The query value or values for the search term. You can use the information in this column to determine how to modify the search to increase or decrease the number of hits.
Total

The Total row appears when you apply selections on the Search Terms page. The Total row displays totals for the search term family. It is not the sum of the individual terms in the search term family. For example, the Total row for the Document and Family columns displays the following:

Document: The number of distinct documents with hits for any term in the search term family, that is, the result a user sees for searches using search term family "has a value" parameters within the specified document set.

Family: The number of distinct documents with hits for any term in the search term family, including all documents in the family, within the specified document set. Standalone documents, which are not a part of a family, count as a single document family. If the specified document set is not family-complete, only the family members that are part of the specified document set contribute to this count.

Counts

Document

The number of distinct documents within the selected document set with hits for the term.

Family

The number of distinct documents within the specified document set with hits for the term, including all of the documents in the family. Standalone documents (not a part of a family) count as a single document family.

Unique document

The number of distinct documents within the selected document set with hits for the term, not including documents with hits for any other terms in the search term family.

Unique family

The number of distinct documents within the specified document set with hits for the term, including all documents in the source/attachment family, but not including documents that are members of a source/attachment family where any document has hits for any other terms in the search term family. Standalone documents (not a part of a family) count as a single document family.

Coding

Recall

Indicates the percentage of documents in a population that are coded as positive by human reviewers’ marks and that have term hits. A higher percentage means fewer positive documents were missed by the term.

For example, if 1,000 documents out of a population of 5,000 are positive, and the term has hits in 850 of the 1,000 documents, the term’s recall is 85 percent.

Precision

Indicates the percentage of the documents with term hits that were coded as positive by human reviewers’ marks. A higher percentage means fewer documents were incorrectly identified as positive (returned as hits) by the term.

For example, if the term has hits in 800 true positive documents, but also has hits in an additional 800 false positive documents, the term's precision is 800 documents out of 1,600, or 50 percent.

Positive with term

The number of documents within the specified document set with a hit for the term and coded with a positive value in the selected coding field.

Negative with term

The number of documents within the specified document set with a hit for the term and coded with a negative value in the selected coding field.

Not coded with term

The number of documents within the specified document set with a hit for the term and not coded with a positive or negative value in the selected coding field.

Positive without term

The number of documents within the specified document set coded with a positive value in the selected coding field, but without a hit for the term.

Negative without term

The number of documents within the specified document set coded with a negative value in the selected coding field, but without a hit for the term.