Historical snapshot. Some original links and images are unavailable; forms are retired.

Information Flow blog

May 11, 2004

Eyes for Text

Categories: Search and Retrieval
After years of explaining "Beyond Search" ideas, I've found no more evocative a phrase than "_Eyes for Text_". Radiating from the phrase are a half dozen ideas. * Humans have _Eyes for Text_. Handed a document, a person can quickly scan the document and extract all kinds of useful information. The key word is _useful_. It's human magic that makes that assessment work in such a broad range of conditions. * Machines have generally lacked _Eyes for Text_. They see the same document as a sequence of bytes, when in fact to us it's really made of words organized in sentences organized in passages, not to mention all kinds of other structure related to conventions, forms, genres, and so on. * Humans are blind in the face of large numbers of documents, while machines are blind to what's in one document. Our human blindness limits our ability to, well, see patterns and anomalies across the whole and connections across elements. While machine blindness means they can't help us much. It would be a case of the blind leading the blind. * _Eyes for Text_ also suggest steps up toward understanding that aren't necessarily all the way up, say to _Brains for Text_. So before we solve all the scientific problems of artificial intelligence and cognitive science, we can master a useful set of primitives that certainly must lay on the path in any case. * Would we be able to understand what's in a kitchen, not to mention navigate around it, or make dinner there, if our eyes didn't pull out useful features like edges, surfaces, corners? * And the edges, corners, and surfaces within documents? They are the entities mentioned, the statements made about them, whether they state relationships, events, or facts. And the sequence of these statements tell us about the topics, authority, applicability, and so on of the text. The first uses of _Eyes for Text_ software engines will still leave humans to perform their magic of knowing what might be useful. However, now aided by machines that can pull out bits of information well enough to deliver them as the human searches or browses, and to create mostly correct, rough sketches of all whats in there. And out there.

Posted by ramana at May 11, 2004 07:22 PM

Trackback Pings

TrackBack URL for this entry:
http://www.ramanarao.com/cgi-bin/mt-tb.cgi/11

Comments