The Enterprise Search "Essay Test"

Andy Moore: As new media assets (e.g., voice, streaming) plus new technology environments (e.g., mobile) start to emerge as business challenges, the temptation is to attack those problems with specifically focused solutions from specifically focused vendors. BUT, as we all know, this kind of thing leads to siloed, non-integrated applications all over the place. Been there, done that.
What can you tell us about your efforts to homogenize AND adapt your solutions to prepare for these specific new technologies (rich media, mobile)?

Jason Hekl: We’ve been working on a mobile interface for the InQuira platform that reorganizes the pages and functions in the application for mobile devices.

This is critically important to field service agents, who may not always have access to a PC-based solution. An agent can configure the application to determine which information displays on his home page. Agents can post questions to a discussion forum, and the application automatically sends response notifications to the agent’s mobile device.

Field service agents are unlikely to adopt and use a knowledgebase solution unless they have both PC and mobile access. They simply will not come back and document on their PC the issues they ran into, and look for solutions. With a mobile application, agents in the field can capture knowledge from forums, and use their mobile devices to recommend content to KB articles. All the elements for effective knowledge capture are at their fingertips, accessible via a mobile device.

Jerome Pesenti: Search is definitely pushing boundaries. In terms of data reach, it’s becoming able to ingest more and more—audio, video, databases, structured data, logs. Everything is actually fair game as long as they are properly and carefully handled. In terms of user reach, search should be ubiquitous and accessible anytime, anywhere and by anybody, helped by the fact that from an end-user standpoint, the search process is as simple as can be, one search box!

In terms of platform environment, Vivisimo continues to leverage platform-independent XML-based standards in a distributed and service-oriented architecture. This allows our platform to be very versatile and grow with our customer needs and evolving requirements. We currently offer a mobile extension to our platform for secure intranet search via mobile devices, can support rich media. We were the first company to incorporate Web 2.0 functionality into the user interface via our own technology.

Vijay Koduri: One of our core principles at Google is "focus on the user." This simple principle guides everything we do, from design to architecture to feature selection to product roadmap. We never try to dictate which technologies or media assets the end-user should adopt. Rather, we closely observe what the end-user is adopting and make sure we support it. It is completely natural for media assets and technology environments to proliferate; that is why our answer is to have a powerful search solution across varied content.

Which leads us to Google’s universal search for business. The whole premise of universal search is that all enterprise content sits behind the Google search box. And the search engine provides relevant results across varied sources of content. On the consumer side, we introduced the concept of universal search in May of 2007, where a single search would search across Web pages, images, videos, books and any other relevant content. The same concept was built into the Google Search Appliance 5.0, introduced in late 2007. Universal search for business enables search across intranets, file shares, databases, content management systems and real-time business data, all through a single Google search box.

Another aspect of universal search is the universality of access. Data can be accessed from any platform, anywhere. For instance a busy sales professional might need to access critical enterprise data from her cell phone. Similarly, a doctor might open up a PDA in between patient visits to access treatment information. Regardless of the technology device, all they need to access is the Google search box, and they can reach all of the enterprise content.

James Waters: A great search company adapts its solutions to what is happening in the market. Multimedia and mobile are two of the hottest areas around. And Coveo has solutions for these. One of the platforms that is arising quickly is our old friend email. It is growing every day as the unofficial file server of the companies of the world. Where is this Word document being delivered to? Email. How will you deal with revisions and version control? Email. And where will you see this email once or twice? On your mobile PDA. So we think that email search and mobile are the two hottest areas around. And let’s not forget audio/video. Training, exit interviews, sales meetings, voicemail. Coveo offers a patented product for searching audio/video files.

Harald Jellum: Our strategy is to make any media asset searchable. So far, we rely on third-party vendors in our deliverables. That is we use independent vendors for speech-to-text processing for audio, and picture recognition software for media. Our efforts are primarily case-driven, and we expect for some time to continue using third party vendors—but this may change if we experience significantly market demand.

Johannes Scholtes: ZyLAB can store any type of electronic information (multimedia, paper, email and other electronic files) on just a normal SAN or NAS in XML structures. If needed, these can be integrated with relational databases, but it is not mandatory. This gives ZyLAB enormous flexibility with respect to setting up various "silos" on blade or file servers. ZyLAB can therefore also integrate easily with other multimedia recording systems.

The ZyIMAGE Analytics Server allows users to automatically enrich their unstructured information with metadata, giving it the structure needed to support organizations’ e-discovery, records management, knowledge management and historical archives programs.

Twenty years ago, only librarians had any sort of structured search tools. Full-text search was a novelty. Today, most Internet search engines and personal search tools only provide a basic full-text search for unstructured searches; all the old school structured access methods are gone. As collections grow larger, the need for a combination of structured and unstructured search techniques becomes more important.