Maryland Bar Bulletin

Publications : Bar Bulletin : November 2010

|

TECHNOLOGY TALK

How do you organize your files? Long, descriptive file names help. Folders are a must, but do you organize them by client, date, subject matter? How do others in your firm organize their files and do you have a standard procedure that everyone uses? Are your files stored on your PC or on the company network? And the biggest question of all – can you find what you are looking for when you need it? In other words, is it working for you?

What Makes Golbalbrain Better?

Globalbrain does not try to derive meaning, nor does it care who, what or why you are searching – it just helps you find what you are looking for as quickly as possible. Globalbrain is extremely successful at handling misspellings, uncommon terms and very large amounts of text in a search query, and results are returned with lightning speed.

Have you worked with optical character recognition (OCR) before? When scanned documents are converted back to text using OCR, spelling errors are very common. As a result, the word you are searching for in a document may never be recognized, even though it is used multiple times because they all contain the same typo.

Globalbrain does things a little differently. Globalbrain employs an associative access or “n-gram” approach to search, which removes the human element, turning words into mathematical representations. How does it do this? Words are parsed into three-letter parts called tri-grams or three-letter components. These tri-grams are then used to index text from the document. For example, the word “sample” would be parsed as “sam”, “amp”, “mpl”, etc. When your search query is parsed in the same manner, Globalbrain compares the three-letter snippets to those in the index. Finally, it rates the possible hits based on matches to those three-letter groups. This is “fuzzy logic” in action. It’s not an exact match on a specific term, but a list of results from which the user can choose relevant items.

What Makes Golbalbrain Faster?

As part of the indexing process, the tri-grams are converted to ones and zeros, making the index footprint very small – usually only 10 percent of the original document repository. For example, if you had a 200GB repository, you would be searching a knowledge base of about 20GB of searchable text.

When you search, your query string simply needs to be converted using the same tri-gram method and quickly matched against the existing data. This is much faster than checking against multi-lingual and multi-variant dictionaries or keywords.

How Do You Search Globalbrain?

Globalbrain finds information everywhere – the personal and enterprise editions support over 250 file formats and locate information no matter where it’s hidden: your PC, enterprise databases, file servers, e-mail servers or the Internet. Through performance and process optimizations, Globalbrain scales from personal laptop configurations to global enterprises, allowing you to search millions of documents and terabytes of data.

How do you find what you are looking for when searching the entire possible range of locations and file types? Well, Globalbrain works best when it’s presented with more information in the query string rather than less. Globalbrain can handle a 20-page search query, which would normally paralyze a typical search solution, with great speed and accuracy.

Multi-Lingual Searching

Because the search query is converted from words to n-grams, your searches are language-independent. Globalbrain looks for patterns and meaning, rather than simply words. Therefore, it delivers accurate results regardless of the language in which documents are written. There is no need to filter for language at any level.

What Else Can You Do with Globalbrain?

Globalbrain’s powerful categorization capabilities set it apart from other search solutions. Globalbrain can automatically sort information into user-defined categories – providing you with a variety of views of your information.

What’s New

The newest edition of Globalbrain, version 5.2, offers some new features:

  • More flexible security for crawled documents – v5.2 can understand permission settings on documents and folders from external applications such as document management systems.
  • Ability to export search results – v5.2 allows users to generate a PDF containing selected search results, download a zip file containing selected documents, or e-mail selected documents to others.
  • Improved display of search results – v5.2 allows users to apply different templates for previewing different document types.
  • “Sticky” query panel attributes – administrators can configure attributes for input fields being displayed on the query panel at any given time, allowing users to limit their search to documents having distinct attributes.

What are Others Saying?

Earlier this year, KMWorld included Globalbrain Enterprise Edition enterprise search solution among the Trend-Setting Products of 2010. The award recognizes clearly identifiable technology breakthroughs that serve the vendors’ full spectrum of constituencies, especially their customers.

“Globalbrain signifies a clear step forward in its ability to search an organization’s vast data repositories with intelligence and speed,” says Hugh McKellar, KMWorld Editor-in-Chief. “Information represents a business’s most vital resource, and the ability to retrieve that resource without limitations of language, misspellings and platform interfacing abilities is an invaluable asset to any dynamic enterprise.”

For more information about Brainware or Globalbrain, visit their website at: www.brainware.com.

previous next
Publications : Bar Bulletin : November 2010

back to top