WebFOCUS Magnify: Wizards, Collections and Security

By Adam Lotrowski

Today’s companies have an incredible number of informational assets dispersed through applications, databases and documents. Immediate access to information is necessary for all resources to effectively perform their jobs. With Magnify, all data can be easily indexed and searched with an intuitive Google-like interface.

But how can prototypes, simple demonstrations, and proofs of concept quickly be created without building an entire search solution? How can each Magnify instance be made accessible from a single environment? How is information secured? Read on for the answers.

Build Quick Demonstrations and Proofs of Concept

Magnify offers a large number of adapters for reaching more sources on numerous platforms through communication methods than any other existing enterprise search platform. With tools for ETL processing, the Magnify platform transforms raw data into search results with mechanisms for quality control, metadata management and record synchronization.

On the other hand, if none of the above is required, index creation can be completed using Magnify’s iWay Wizard. With a sample set of data (in Excel, SQL Server 2005 or HSQL) a preliminary Magnify solution can be configured in less than an hour.

First, the sample set of data must be configured for Magnify’s iWay Wizard. This requires molding the data into a search result and thereby making it ready for automatic indexing. This can be as simple as concatenating fields together and changing column titles.

There are six parts that define a search result:

The dataset’s field names are used to instruct how Magnify’s iWay Wizard handles each field when transforming a record into a search result. A dataset can be created with an Excel formatted WebFOCUS procedure or by creating views in a database. As long as the dataset uses documented field names and data that make sense when fashioned as a search result, it can be fed to Magnify with easily modifiable, ready-to-go iWay templates.

In pseudo-FOCUS code:

See Screen 1 on the next page for an Excel file with a configured dataset ready to be fed to Magnify via the iWay Wizard. The Magnify search presents the Excel records converted to search results. The title’s main link executes a search on Google.

Organizing Search

People trust a Google search and they understand the simplicity of its single input box. Magnify’s Google-like interface is just as user-friendly. In the enterprise, however, relevance and the ability to interact with search results and filter them down to the most useful document become even more important. It’s possible to assign parts of a search result’s content to defined meta-tags, instruct how dynamic categorizations trees are automatically created, and make values available for result analysis and sharing. It’s also possible to dynamically customize accessibility to search results through security and perform drill-downs for additional reporting.



Screen 1

New to WebFOCUS 7.6.9’s Magnify is Collections, which provide administrators with a way to organize search results into specific groups. This allows users to select a collection, which narrows their search to a specific part of the index, prior to their submission. Collections can be a virtual subset within a Lucene index, a Lucene index stored in its own dedicated folder, external to or found in the Lucene directory, and various combinations of virtual subsets and indexes residing across the file directory.

Screen 2 shows a search for the term “Magnify” in the Marketing collection and that the Focal Point collection is the next one to be searched. This example is in Information Builders’ Electronic Content Library. Each department’s content is fed to Lucene indexes stored in their own folders in the default Lucene index directory. Multiple-level sub-directories found in the defined Lucene index directory are implicitly (or automatically) included in the Magnify search and only require it be presented via Magnify’s Style Sheet.



Screen 2

Configuring an implicit Magnify collection is easy to do.

1. Verify that your Lucene index directory has been backed up and empty, except for the two collection files (path defined in the WebFOCUS Administration console).
2. Create a new folder with a short, meaningful name.
3. Create a new Lucene index and move it to the folder created in Step 2.
4. Open the Magnify Style Sheet in a text editor (path defined in the WebFOCUS Administration console).
5. Locate attribute show_collections and change value to 1 (Line 83 in default version).
6. Locate attribute collections_descriptions and, in the desired position, add the collection name, visible to users. The syntax is a comma-separated list and must end with a comma (Line 61 in default version).
7. Locate attribute collections_values and, in the desired position, add the exact folder name containing the index files dedicated for the collection in Step 6. The syntax is a comma-separated list and must end a comma (Line 62 in default version).
8. Save the Style Sheet, restart Tomcat, and perform a search (Magnify link found on WebFOCUS Welcome page).
9. Repeat as many times as you like, but as a best practice, keep a short and meaningful list of collections:

The Magnify Style Sheet with collections turned on and two collections defined. The Default Collection is the index found in the root Lucene directory. Newsletter Sample is the collection implicitly created and used in the Magnify iWay Wizard section above.

All other collections are explicit and require modifying the collection files and the style sheet (as described). In the future, explicit collections will be configured for conditional collections, rule-based collections, style sheet assignments, localization settings and secure collection access.

Powerful Security

As powerful as search is for finding information, security is needed to ensure that only authorized users benefit from the ease of access to information:

■ Single sign-on integration. Magnify integrates with existing single sign-on frameworks. Authentication passes to Magnify and is used to authorize user access to information.

■ Multiple credential support. Different credentials are required for authentication when indexing multiple sources of information. After you submit your search, you are prompted for each required set. Magnify dynamically generates a single logon for each user name and password required.

■ Hide entire results. If you do not have access to a search result, it will be hidden from you entirely. There are no indications anywhere in the application to indicate that the search result exists. Without authorization, Magnify completely restricts results from being seen.

■ Hide parts of results. Some security rules will allow you to see a search result, but only authorize specific information. For example, if you filter search results by salary range but are not allowed to see your manager’s salary, and were to click the range matching your manager’s salary, you will not find your manager in the results. If you filter on title, which you are allowed to see, you will find your manager in the results. Magnify security allows for hiding parts of search results including the metadata.

■ Present alternate-result rendering. Sometimes you may be restricted from viewing a search result, but you are still made aware that it exists. Instead of seeing a search result, you are presented with a message to log on, sign up for an account, subscribe, or pay a fee in order to view the hidden result. Magnify allows for search results to be modified dynamically at runtime based on a combination of the user profile and information returned.

■ Security API. Magnify’s security is flexible enough to work with almost any security model. The Security API provides the framework to authenticate and authorize user access. The Security API allows access to HTTP requests, issues alarms on suspicious activity, and sends alerts on erroneous behavior. The Security API is also applied while information is being fed to the Magnify index.

In my next article, I will introduce feeding files to Magnify search. See you then.