What to do after installation

Once Semaphore has been installed, initially configured, and tested, then you will need to do the following:

1. Create and/or update the model: This involves augmenting the model for use with the Semaphore software (including optimization for automatic classification purposes, if required). The details of this are outside the scope of this document.

2. Implement functionality on your systems: This involves implementing specific interfaces to expose model information to your users. The requirements for implementing Semaphore functionality is described in the following sections.

Note: The following sections assume that you will need to implement custom functionality for your systems to integrate Semaphore functionality. If you are using Semaphore-provided integrations, refer to Integrations

Automatic classification functionality

If you would like to have your system automatically classify documents whenever they are added to your system, then you will need to submit those documents (either in batch or as documents are added/modified) to the Classification Server for classification. Refer to the Classification Server - Developer's Guide for information on the programmatic interface that can be used to submit content to Classification Server.

Additionally, you may want to save the classification information returned from Classification Server to use it for a search interface. For example, you could display only terms returned from Semantic Enhancement Server for which documents have been classified (it does not make sense to display terms for users to search on if those terms have no content).

Model enhanced search functionality

Adding a model to your search experience can greatly improve the ability of users to find the information they are searching for on your site. There are a number of elements of search functionality that you could expose. For example:

• Concept Mapping: This functionality of Semantic Enhancement Server, for given search text, returns a list of concepts that it believes are relevant for the search. In addition, you can restrict the terms returned to only those that have content in your system (see the "Automatic classification functionality" section).

• Topic Maps: Whenever a search is performed, you can display a list of model terms that have been found present in the result set (assuming you have stored the classification information relevant for each document). This information can then be used to filter the search results appropriately.

• Model Information: Whenever a term is selected, you can display information relevant to that term that is stored in the model. For example, meta information could include a URL to a picture, descriptive text, related information, etc.

• A-Z: An A-Z-style page contains a list of key terms from the model that the user can then browse to either perform a search (when they click on a term a search is performed to return only documents classified against that term) or simply to display model information.

To learn how to leverage this functionality with your search solution, refer to the Semaphore Semantic Enhancement Server (SES) API Reference, relevant Semaphore deployment, and integration guides.