Indexing Methods can lead to search frustration if their functionality is not fully understood. In addition to the Stop List, Indexing Methods applied to profile fields are top reasons why the search results do not seem to match the document set you are searching. The leading cause is the default indexing method applied to the Full Text profile field of Alchemy databases.
The Full Text profile field has a default indexing method of All words excluding pure numbers. The Full Text field is searched when you are searching the content of your database as opposed to a profile field displayed in the profile pane of the Alchemy desktop client interface. Because the indexing method is defined as All words excluding pure numbers, you will not be able to search and retrieve results based on numeric data alone. The pure numbers of the content have been excluded with this particular indexing method.
There are some related articles that discuss Indexing Methods in greater detail. Following are two such articles that will help you better understand Indexing Methods within an Alchemy database:
http://solutionbench.com/2016/02/22/how-do-i-define-an-indexing-method-in-alchemy
For this article, we will walk you through the steps necessary to change an Indexing Method for an existing profile field using the Full Text field as our example.
Select a Database using Alchemy Administrator
Using Alchemy Administrator, Select (1) the database you wish to modify in the Tree View display of the desktop client. Any changes we make in the Design menu of Alchemy will only impact the currently selected database.
Display the Fields Dialog from the Design Menu
From the Design menu, select Fields…
Select a Field to Edit
Select (1) a field to edit and Click (2) the Edit button to edit an existing profile field in Alchemy. In this example we are going to modify the Full Text field of an Alchemy database. This is the field used by Alchemy to store the content of documents and includes the search full page OCR text of TIFF images.
Change the Indexing Method
Change the indexing method from the default of All words excluding pure numbers (1) to Default DataGrabber (2) and Click OK (3) to change the indexing method to Default DataGrabber. The Default DataGrabber indexing method is the best standard indexing method at indexing financial data with currency symbols along with thousands and decimal separators.
This change is a point forward change, meaning all previously indexed documents remain indexed with the old indexing method. Only newly added documents will be indexed with the new indexing method unless you Clear the database index and rebuild the database. If this is done, all documents will be reindexed with the new indexing method selected above.
Clear the Database Index
With your database still selected in the tree, select File > Clear Database…
ONLY Clear the Index
The only completely restorable option in the Clear Database dialog is the Clear index option. Selecting this option will require you to Build the database again to regenerate the searchable index. This process can take some time depending on the size and number of documents and profile fields in the database, so caution should be taken as to when this procedure is executed. While the database is being rebuilt, no additions to the database can occur, so this procedure may be best scheduled after normal production hours making sure it has time to complete before any backup procedures take place.
When you are ready to clear the index of the database, Select (1) ONLY the Clear index option and Click OK (2).
You can learn more about the Clear Database dialog in the following article:
http://solutionbench.com/2016/02/20/how-to-clear-a-built-alchemy-database
Confirm your Selection
Click Yes (1) to confirm your intent to clear the database index.
Build the Database Again
Click (1) the Build Database icon on the toolbar and Click OK (2) after confirming your build settings to build the database and regenerate the previously cleared index.
Verify Search Contents
You should now find that financial data and pure numbers exist in the content index for your database when you open the Search dialog, select the Document Contents (1) tab for a Specific Database (2) and click the Lookup Table (3) drop down button.
Recent Comments