Products: Enterprise Search Engine

Enterprise search has always been a weak link in connecting people, information and insights inside organizations. People have become spoiled by the quality of Internet search and expect the same clairvoyance in enterprise search. But they are usually disappointed.

We believe that enterprise search can be significantly improved, not by making it more like Internet search, but by making it better adapted to the distinct requirements of the enterprise search context. Unlike Internet search, enterprise search is aimed at a relatively small corpus (millions rather than billions of search objects), but encompasses a wider range of object types – not just web pages, but documents in multiple formats, applications, datasets, blogs, communities, etc. Depth of search, rather than breadth, tends to be the primary driver of success.

Enterprise search also needs to be better than Internet search at integrating people finding, or expert finder functionality, with content search. In an organization, people are paid to contribute their knowledge and expertise to the good of the enterprise. Search must help people find experts who can help them win sales, design great products, launch winning marketing campaigns, or otherwise contribute to organizational success.

In addition, we have found in our extensive research with corporate searchers that search inside an enterprise is inherently multi-faceted. Findability in enterprise search is greatly enhanced by the search engine’s ability to leverage and expose multiple “facets” or dimensions of classification. For example, zeroing in on the right result occurs much faster and more accurately when an initial result can be filtered by business process relevance, product relevance, location, organization, language, document type, and other dimensions. In this regard, enterprise search is more like finding a particular kind of coffee maker on Shopping.com than it is like finding information about Galileo on the Internet.

Finally, we have learned that search is a much more collaborative and repetitive activity inside the enterprise. Knowledge workers share searches and search results. They find value in saving searches, running them again periodically, and identifying if and how the result has changed. They often use search to get the same results that an RSS feed would provide much more intuitively.

  SearchScience Research Highlights

We have studied and interviewed searchers in a wide variety of companies and governmental agencies. Here are some representative quotes that capture what we have heard:

• 80 percent of the time, intranet search results do not return what I’m looking for.
• I hate scrolling down through long search result lists only to find there is nothing there I’m interested in.
• Searching for experts if you don’t know their names is essentially impossible, but would be incredibly valuable if there were a way to do it.
• I’m not the first person to look for this information, which can’t I leverage what others have done?
• I often do the same search every day and would like to automate that task.
• I would like to see my search results grouped into meaningful sub-categories – for example, by business process, document type, products referenced, or projects referenced
• When I finally find what I was looking for, I would like to share my successful search with my co-workers.
• When I submit content to the intranet, I want to be able to tag it with consistent categories so that others can actually find what I have done.

  Differentiating Features

The SearchScience Enterprise Search Engine has been designed from the ground up to provide a differentiated set of features to support searching and findability inside an enterprise:

• Enterprise taxonomy-enabled searching – ability to search for content or people by business process relevance, location, organization, community, product, project, or other identifying business category
• Personalized and augmented search based on the searcher’s user profile – location in the organization, job type, business process participation, etc.
• Expert Finder searching – finding experts based on their user profiles, community membership, blogging activity and other online behavior
• Enterprise search administration – ability manage all enterprise content indexing, tagging, and search logging from one central administrative console
• Faceted search – ability to filter search results based on multiple facets
• Automatic taxonomy-driven, post-query categorization of search results
• Ability to save searches and re-run them on a periodic schedule
• Saved Search deletion integrated into corporate document deletion policies
• Search collaboration – ability to share search definitions and results with co-workers
• Seamless integration with content security tagging – all content security tags transparently enforced at the point of search based on searcher access privileges
• Mobile-enabled search and retrieval
• Support for “guest” logins to enable anonymous searching (can be switched on or off by enterprise administration)
• Built-in rating and feedback system – ability to rate searches and results (customizable)

  Engine Technical Specifications

Search Science applications employ a modular and extensible architecture to support secure, scalable and reliable services. The Search Science Engines are designed to integrate smoothly into existing IT infrastructure.

• Integrates with existing Security IT LDAP/Active Directory Infrastructure
• Can be rapidly deployed on easy to use off the shelf hardware
• Easy to use and maintain
• Deployed on Industry standard J2EE Application Servers
• Supports Integration with Verity Enterprise Search Engine using custom adapter
• Supports indexing of Microsoft Office, PDF and standard email formats
• Supports indexing of SCM, ERP and CRM system data via custom adapters

  Hardware Requirements

PC with 1 Ghz or higher processor clock speed recommended; 800MHz minimum required (single or dual processor system); Intel Pentium/Celeron family, or AMD K6/Athlon/Duron family, or compatible processor recommended

• 1 gigabytes (Gb) of RAM or higher recommended
  (512 MB minimum supported; may limit performance and some features)
• 1.5 gigabytes (Gb) of available hard disk space
• CD-ROM or DVD drive

  Software Requirements

OS	:	Microsoft Windows 2000 Windows Advanced Windows 2000 Windows XP Pro
Database	:	Microsoft SQL Server 2000 Oracle 8i/9i
Application Server	:	JBoss 3.x/4.x WebLogic 8.x/9.x
Java	:	Sun JRE 1.4.x