Inside Search:
What "Search" Really Means and Where It's Going

Old search is out. It’s no longer just about finding information on the Internet and on desktops. It’s certainly not about secret algorithms and how to optimize one’s search rankings. And it’s not only confined to basic keyword approaches.

What's next: A much broader, more comprehensive definition of search. Search is now really about finding information from any source (books, e-mails, newspapers, databases, reports, documents, websites, blogs, podcasts, videos), at any location (online or offline, networked or local), and in any form (unstructured or structured, digital or analog, in multiple languages). And then it’s about putting everything together and “making sense” of it all.

From searching to sensemaking

Sensemaking ("making sense" of it all) is necessary for decision-making: whether it’s combing the blogosphere and other media to monitor a company’s brand reputation, or sifting through large, heterogeneous content collections to identify potential terrorist activity. However, as today’s information environments become increasingly complex, decision-making has become much more difficult and time-consuming.

So understanding massive and diverse content is not just a simple matter of consuming information or finding it faster. Instead, sensemaking involves multiple tasks — often handled by various individuals working together — such as searching, filtering, and extracting relevant content; rapidly reading and absorbing meaning; "connecting the dots" or detecting patterns; digging deeper and uncovering hidden relationships; gathering appropriate evidence and summarizing results; and efficiently organizing and presenting information.

To successfully accomplish these tasks, people need semi-automated tools and methods that help: 1) focus attention and budget resources; 2) determine where to apply different levels of analytic depth; and 3) avoid cognitive biases or common decision-making traps (especially in mission-critical situations).

Bottom line: more advanced approaches for interacting with information are needed, and simple keyword search won’t do the job.

PARC researchers searching for meaning

PARC’s approach to sensemaking combines language analysis/text mining technologies on the back end, with user-interface and information visualization technologies on the front end. Also involved are intelligent document recognition research that takes advantage of visual structure in document images, and security research that redacts mined content when necessary to protect individual privacy. Together these technologies help people efficiently yet deeply integrate, organize, analyze, and display massive and diverse information.

PARC’s influential natural language researchers — who invented the technology behind most spellcheckers today — have constructed a “semantic pipeline” of technologies, from extracting facts and representing knowledge to inferring beyond the text. PARC’s pioneering user-interface research group, which yielded the GUI among other things, has shifted the focus from human-computer interaction to human-information interaction (emphasizing how users interact with information not just devices).

The result: dynamic, user-sensitive technologies that enable deeper understanding, better decision-making, improved communication and collaboration, and greater productivity — moving beyond basic search and towards sensemaking.

Interested in learning more?

• www.parc.com/sensemaking
• Contact Lawrence Lee, Director of Business Development, Intelligent Systems Laboratory
• Contact Linda Jacobson, Manager, Marketing & Communications, pr@parc.com or 650-812-4035