EnterpriseSearchCenter.com Home
  News   Features   White Papers   Research Reports   Web Events   Conferences  
 
RESOURCES FOR EVALUATING ENTERPRISE SEARCH TECHNOLOGIES
February 20, 2008

Table of Contents

Taming Multiple Search Engines in Your Organization
Reuters Releases Open API For New Calais Web Service
Fortiva Announces Low-Cost Email Archiving Solution
Improving content monetization
Fourth-generation KM
Collexis Acquires Lawriter LLC
Coveo and NavigationArts partner
Serious e-discovery
Locating housing for disaster victims

Taming Multiple Search Engines in Your Organization

One-size-fits-all enterprise search is dead—if it ever existed. At the same time, search that makes an organization’s assets more readily accessible has become a critical mission. Customers, partners, and employees all expect to be able to "find answers" via search technologies.

Thus, organizations increasingly deploy search engines to help satisfy these expectations. These engines are typically called site search, department search, intranet search, or application search, rather than enterprise search.  However, the fact is that organizations that have one search application usually have several. A 2007 enterprise search survey I did for the publisher of the Enterprise Search Sourcebook, Information Today, Inc., and my employer, Shore Communications, found that 62% of respondents had more than one search solution in place. Noted industry analyst Steve Arnold reports that the typical Fortune 500 company uses solutions by at least five search vendors.

Download the free PDF


This proliferation (and multiplicity) may not be obvious, even to those within the companies deploying the solutions, since responsibility for search strategy is frequently undefined within the organization. One reason for this is because search doesn’t exist independent of the content that is being accessed. It creeps in with add-ons to content management systems (CMSs), business intelligence (BI) applications, records management, document management, or knowledge management systems. In addition, ecommerce systems with baked-in search capabilities can add to the growing number of search vendors used by an organization. As an independent application, search can be tuned to specific business-process needs, which means it comes in many guises.

Thus, many companies face more than the already challenging issue of managing a single search engine: As the number of solutions deployed within an organization increases, distinct issues emerge in reconciling these disparate solutions. The major problem is that each search product builds indexes to the existing content using a "secret sauce" that’s incompatible with other vendor indexes.Thus, finding answers across different application repositories, each with its own index, is not a straightforward proposition.

Strategic Planning for Search

One enterprise search myth is that all information in the organization should be accessible via search. Organizational content, however, is more difficult than the relatively simple world of consumer search on the open web, which is primarily HTML webpages and unstructured content. Information created and controlled by the organization is complex, from the content perspective and the technology perspective.

There are some facts to consider about content and how it relates to search functions:

The value of content within the organization varies. Indexing a server that contains years of cafeteria menus is a waste of resources, creating yet more information overload. Indexing technical reports representing the intellectual capital of years of R&D, thus enabling idea discovery, has potentially high ROI. Improved findability for ecommerce companies, particularly those with large product catalogs, can result in increased sales.

Job functions require different content. A financial analyst may need current sales by product line, even down to the individual part number. Customers and partners want data sheets with technical specifications. Defining subsets of content for job functions is the domain of management, not technology.

Security complicates search. Privacy of employee records is essential, though employee benefit information should be made available to everyone. Access to customer records may prove useful for certain job functions, but it may be regulated by federal and state laws. Access to trade secrets is typically restricted to those with a need to know. Realtime identification of fraudulent transactions may be mission critical. Thus, access and security policies need to be in place before unleashing the power of search technology.

Technology Planning for Multiple Vendors

Another myth of enterprise search is the feasibility of standardizing on a single search vendor. Search engine software creates proprietary indexes and relevancy ranking, with each having different strengths and weaknesses. Products from the same company, say IBM or Autonomy, do not necessarily provide compatible upgrade paths as search applications grow. There are some simple facts every company must face when attempting to reconcile search solutions:

Legacy systems are part of the search landscape. Many business intelligence and content management vendors integrate search into their product suites with OEM relationships with enterprise search companies. This ensures best-of-breed functionality within the suite, but not necessarily between vendor suites.

Mergers and acquisitions play havoc with standardization. Acquiring a company is based on business fundamentals, not compatibility of software systems. Provided the existing systems work well at the application level, there is little ROI in switching search software. However, the consolidation of software as search companies are acquired by major players can result in product "orphans," leaving your organization without support.  Federated search needs to be part of the technology plan. Federated search refers to the capability to search multiple indexes without creating yet another index. Also called meta-, blended, or universal search, this capability can be implemented in various ways: realtime federated searching using on-the-fly queries, portals, webpage mashups, or structured top-50 queries. Implementation is highly dependent on access and security requirements, which can vary with each of the underlying content repositories.

Care and Feeding of Indexes

Search engine indexes and the information retrieved from enterprise repositories rely on the accuracy of the underlying data. Maintaining integrity of the applications that create content is crucial for business decision making, yet organizations typically underestimate the time and effort it takes to maintain clean and relevant content repositories.

Here are some issues that must be considered: Records management is an underappreciated aspect of search. Regulatory compliance and ordinary retention schedules should be part of managing business risk. Indexes exist separately from the content, so they have their own update schedule. Overnight or weekly—rather than real-time—updates can make search results appear out-of-date. Search technology makes misspelled, noncompliant, and dirty data glaringly obvious.

Indexes grow as the numbers of records applications grow. Index files vary in size but can be an additional 50% to 200% larger than the size of the primary data files, depending on the expansion factors. Scalability of the search solution should be a major consideration in selection and implementation.

Relevancy of search results depends on context, not popularity. The purchasing department may work with a dozen "business card" vendors worldwide. A "business card" query for most individuals in an organization, however, means a short how-to procedure for ordering new business cards. Search technology alone doesn’t solve the job of interpreting the query to provide relevant answers.

Managing expectations is a major challenge for search implementations within the enterprise. Search is a tool to provide answers to problems. Finding those answers across enterprise repositories requires a higher level of relevance than open web consumer search, yet the underlying content structure is more complex and access is more constrained. Just as there are no one-size-fits-all answers to organizational questions, enterprise search is not one-size-fits-all, and most organizations will need to manage multiple solutions to supply the different answers individuals seek.


About the Author

JEAN BEDORD (jean@bedord.com) is a findability and search consultant with EContent Strategies, senior analyst at Shore Communications, Inc., and faculty lecturer at San Jose State University. Her report on "Enterprise Search Deployment, Usage, and Trends" is available right here on the Enterprise Search Center.  Click the link above or type in this URL: www.enterprisesearchcenter.com/Reports/Details.aspx?ResearchReportsID=38.


Back to Contents...

Reuters Releases Open API For New Calais Web Service

Reuters announced the release of an open Application Programming Interface (API) for its new Calais web service. The API, which provides commercial and non-commercial parties with free access to semantic tagging capabilities to enhance their applications and content, is available at OpenCalais.com. The Calais web service enables publishers, bloggers, and sites of all kinds to automatically metatag the people, places, facts, and events in their content to increase its search relevance and accessibility on the web. It also lets content consumers, such as search engines, news portals, bookmarking services, and RSS readers, submit content for automatic semantic metatagging. Calais is a new Reuters initiative that supports the interoperability of content and the development of the semantic web.

(www.opencalais.com, www.reuters.com)

Back to Contents...

Fortiva Announces Low-Cost Email Archiving Solution

Fortiva Inc., a provider of on-demand email archiving, announced the availability of the Fortiva SmartStore archive, a cost-effective way for businesses to meet the demands of the Federal Rules of Civil Procedure (FRCP), without adding to their existing infrastructure or IT staffing requirements. This Software as a Service (SaaS) solution allows businesses to centrally archive all email, enforce policies and litigation holds, perform enterprise search, and easily conduct early case assessment, all for the same or less than the cost of storing and managing the data on enterprise storage in-house.

(www.fortiva.com)  

Back to Contents...

Improving content monetization

MuseGlobal has released two new products, News Hound and Blog Hound, which, the company says, have been designed to aggregate, integrate, enrich and deliver comprehensive, current and relevant content to customers.

News Hound and Blog Hound fuse disparate content sources in both enterprise and consumer markets. Available through OEM and distribution partners around the globe, News Hound and Blog Hound reportedly will also help online news, media and entertainment companies to better monetize their content.

MuseGlobal claims News Hound and Blog Hound allow organizations to:

  • create custom feeds from both structured and unstructured news sources, including search engines, subscription databases, media outlets and Web mining applications;
  • access up-to-date content instantly with query-based searches and alert feeds;
  • migrate and deliver information into third-party or custom applications; and
  • generate new revenue streams from company-owned and syndicated content.
News Hound and Blog Hound enable real-time analysis and results rankings, integrated results from multiple search engines and a single, centralized interface and administrative system. MuseGlobal further says its core technologies integrate with any platform and enable simple installation and administration of connections to content services.

Back to Contents...

Fourth-generation KM

Thomson West has introduced West km 4.0, which it describes as a knowledge management application that lets legal professionals capture and maximize their expertise and intellectual assets.

The company reports West km 4.0 features a new technology platform plus enhanced search and classification functionality, giving users, it claims, a powerful competitive edge by making institutional knowledge and expertise broadly available to legal professionals within an organization. West km 4.0 allows users to:

  • extract the key metadata directly from the text of litigation documents that includes actual document title, document type, high-level jurisdiction and court;
  • view documents with the title as well as content portions, or "snippets" to gauge the document value;
  • import additional metadata for more complete document retrieval results; and
  • receive frameless delivery of results and documents for improved intranet and portal integration.

Back to Contents...

Collexis Acquires Lawriter LLC

Collexis Holdings, Inc., a developer of high definition search and knowledge discovery software, announced that the company has acquired Lawriter LLC, which owns legal research service Casemaker. Total consideration for the acquisition was $9 million, including cash, common stock at $0.75 per share and future financial obligations, plus an earnout arrangement. Lawriter is a legal online provider to the small and medium law firm market that comprises 450,000 attorneys.

(www.collexis.com)  

Back to Contents...

Coveo and NavigationArts partner

Coveo and professional services firm NavigationArts have announced a strategic partnership. The two companies will collaborate to deliver platform-class search solutions that improve the overall user experience, while connecting users to the information they need.

NavigationArts provides professional services to help clients plan, evaluate, design and develop effective Web sites, intranets, portals and rich Internet applications. NavigationArts reports it creates online user experiences that drive business value for communications and commerce.

Back to Contents...

Serious e-discovery

Vivisimo and Wave Software have joined forces to provide a combined solution that is said to offer highly productive first pass culling and pre-review for electronic discovery and regulatory response.

With Vivisimo’s product, Velocity 6.0, quick concept search and culling can occur before attorney review. Velocity’s Enterprise Search platform allows for conceptual and metadata search, annotation and tagging of documents, all of which result in more productive culling, the company says.

Wave Software’s product Trident Pro is designed for first pass data culling by fast deduplication and keyword filtering. Any user can rapidly cull and create a pristine PST or NSF without duplicates, and with all metadata intact. Trident Pro provides reports and logs that outline what was processed. Even with multiple PST or NSF files on the source data, Trident can remove duplicate e-mails and create one resulting PST or NSF file per custodian.

Back to Contents...

Locating housing for disaster victims

The U.S. Dept. of Housing and Urban Development (HUD) has developed a Web–based system to help federal, state and local agencies find housing for people displaced from their homes by natural disasters.

Citizant, a provider of government professional technical solutions, has worked with HUD on the National Housing Locator (NHL) system, which uses Web 2.0 concepts and a data-centric, service-oriented architecture (SOA).

Following Hurricanes Katrina and Rita, the government realized it lacked a national inventory of available housing and embarked on the NHL project to fill that void. The inter-governmental tool combines federal housing resources and commercial housing locators in one platform. It enables rapid identification of available housing based on various search options, according to a press release from Citizant.

Citizant also reports that it has developed an integrated customer service/case management system to handle evacuee placement and emergency assistance delivery. The system has geospatial planning capabilities that help officials understand and compare options, and users can set up what-if scenarios to find the number of available units within certain distances. Web service capabilities allow FEMA and the U.S. Dept. of Veterans Affairs to integrate the NHL system into their online systems, according to the news release.

Citizant further reports that the team started developing the NHL in December 2006 and launched the first release in January 2007, and that the system has already helped relocate families displaced by tornadoes, floods and fires.

Back to Contents...
 
[Newsletters] [Home]

Problems with this site? Please contact the webmaster. | About ITI | Privacy Policy