Text Retrieval Engine is Crucial to A Web-based Document Repository

The evolving technologies of the Internet, along with the demands of the litigation process, were the driving factors behind the development of InfoDox.com, IE Discovery's Web-based document repository. As IE Discovery developed, offering off-the-shelf litigation and document management applications and designing custom databases, the concept of a comprehensive Web-based document repository began to evolve. To meet the varied demands of its legal clients, an efficient interface to the Internet was crucial.

To insure that InfoDox.com met its design criteria, IE Discovery needed a search capability that was fast, secure and flexible enough to hold warehouse-sized volumes of documents yet sufficiently cost-effective to support smaller document projects. It was also important that the systems have powerful searching capabilities, be continuously up-to-date, have Internet accessibility, but not burden the user with a tremendous amount of hardware and software maintenance.

The development team realised that the ability to quickly sift through very large collections of textual and fielded document data and deliver document images over the Internet was fundamental. IE Discovery's primary sources of documents are litigation documents, or documents related to legal issues. These "cases" tend to evolve over time and a solution would need to be flexible enough to adapt to changing circumstances. For example, if a case is related to injuries persons received due to a company's alleged negligence with its chemical waste products, it may be months or even years until all medical records are obtained from all parties involved. The document collection size could double or more in size. In contrast, a client may want to have all their case documents loaded into a database in order to weed through and remove any documents that are not important to the case. In this example, the database size could reduce drastically from where it began.

Because these document collections can be massive, InfoDox.com also had to provide the ability to jump directly to the specified text and image providing both a view of the documents page and taking the viewer directly to line of text searched upon in the retrieved files. Document collections are collected from numerous sources. Because these documents are all received in different file formats, some hard copy, others may be taped video depositions, the system needed to receive files from a wide variety of sources. Handling multiple file types was essential, as was the need to convert these files to formats that could be displayed and searched from the Web. Through months of research, the IE Discovery development team was able to meet these product requirements as well as deliver efficient image delivery over the Web.

Each InfoDox.com database consists of a document and related "names mentioned" table. For each new database loaded into InfoDox.com, the customer is asked to answer the following questions concerning their data needs:

1) What properties (fields) of the documents do you want available to capture? Examples: Document Date, Document Type, Bates Number, etc.

2) Which of the above fields should pick-lists be made available for searching and entering data to ensure consistency in the data?

3) What values should be available for each of the pick-lists described in step 2?

Once the customers requirements have been determined, a database administrator at IE Discovery will design the appropriate database structure and build the database as part of the InfoDox.com service. IE Discovery takes data, or documents in this case, and breaks them down into fielded data (data entry either within InfoDox or loaded from a data entry vendor), images (scanned hard copies of the documents) and OCR (full text from the images). All the converted data is then loaded into the InfoDox application. All document processing, imaging, coding, OCR, etc., is provided by IE Discovery if needed. At this stage InfoDox.com utilises the dtSearch Full Text Retrieval Engine and a relational database search engine resulting in delivery via the users Web browser using standard HTML. The user receives images that are the result of combining a specialised high-compression scheme with streaming text which is converted to HTML with hit highlighting and fielded data.

dtSearch's Text Retrieval Engine provided all of InfoDox.com's search needs including highlighted results in full text searches. By combining their search engine and advanced Internet imaging technology, the specified speed and performance requirements were achieved. dtSearch's powerful search engine, rich feature sets, and professional technical support staff allowed IE Discovery to develop InfoDox.com in a fraction of the time normally required for a project of this scope. The final product allows clients to search mountains of textual data instantly while incorporating regular expression, phrase, Boolean, stemming and thesaurical searching capability. Users can even take advantage of the advanced fuzzy searching capabilities of the engine to find documents even if the terms they seek are misspelled due to typographical or OCR interpretation errors.

To ensure that InfoDox.com documents remain secure, IE Discovery incorporated a protocol called Secure Sockets Layer (SSL) to identify authorised users, prohibit unauthorised access, and encrypt data transmissions from the document repository to the user's Web browser. This is the same high-end technology used by large financial institutions to transmit their extremely sensitive data and the Internet standard for processing credit card transactions.

With the introduction of InfoDox.com, IE Discovery, Inc. pioneered the use of the Application Service Provider (ASP) model to the field of document imaging and litigation support. While InfoDox.com's success is due to several technologies that IE Discovery combined to produce a powerful, yet cost-effective solution for their clients, the performance of the dtSearch Text Retrieval Engine was an important factor. "The product allows InfoDox.com users to search millions of documents in seconds instead of minutes or hours."

For further information, please call IE Discovery at +1 (512) 833-5588, e-mail: sales@iediscovery.com, or visit IE Discovery online at www.iediscovery.com