connecteddale

The strategy conversation you can only have here

The human approach to finding information on the internet

1997-01-21

If hunting for information on search engines has left you a little frustrated then perhaps you should make use of one of the newest search techniques on the web - the human being.

Five undergraduate students from the University of Rhode Island have started a project called HumanSearch which accepts, enquires and returns answers in everyday English.

Unlike search engines that you may have become accustomed to, HumanSearch doesn't ask you to enter queries using any syntax other than what you learnt at school.

Questions such as "What is the population of Portugal?" or "When does it rain in Brazil?" are answered by students who collect the answers on the internet and in the school library and then provide an easily understandable response by e-mail.

HumanSearch could be seen as a human interface to search engines. Let's face it, search engines aren't always easy to use. Without going into any advanced or customised searches, I tried the first question above on some of the more popular search engines.

AltaVista returned 200 000 documents, HotBot didn't find anything, Lycos found "Whatcom County Population" amongst 52566 other documents and Excite found 678 311 documents.

Both AltaVista and Excite returned a document containing the Portuguese population in 1989 as their number one document. So as a first attempt, I didn't get much gratification from the search engines.

If I were to dig a little deeper then I am confident that I could find the answer. But why dig around if someone else will do it for you? And that is exactly what HumanSearch will do for you at no cost to yourself.

The downside to HumanSearch is that they can take up to 48 hours to return an answer. This requires you to think your questions ahead of time. It may not satisfy the information junkies who require instant results but is likely to become invaluable to people doing research projects.

HumanSearch was established when its founders experienced first hand the problems associated with search engines. Information overload is one of the most common challenges. If your search is not specific enough then it is common to have hundreds of thousands of documents returned to you. To make your search more specific requires learning syntax that is often based on computer logic. This is acceptable to computer geeks but not to the man in the street.

Another company offering a similar service is answers.com, who run their search engine as a commercial venture. Users get charged depending on the complexity of their questions. Prices range from US$1.79 for an easy question to US$11.99 for a hard question. If your question falls out of the scope of their categories then answers.com will provide a quote before starting the search.

HumanSearch and answers.com add a new category to the existing two types of search engines. AltaVista, Excite, HotBot and Yahoo make use of digital robots that crawl around the world wide web collecting information about all of the pages that they find.

Yahoo, on the other hand provides a structured directory of information organised into categories and sub-categories. It is updated by humans but you still need to use computer syntax to answer your queries.

Predictions are widespread that technology and more specifically the internet will cause massive job losses as people are replaced with computers. Search engines could be a good example of technology making a task more efficient but requiring humans (performing a different, more advanced task) to make them accurate.

Just as scribes were made redundant and new opportunities for authors opened up in 1450 when Johann Guttenburg invented the printing press, so too might search engines make librarians redundant but open up tremendous opportunities for information specialists who are able to separate the knowledge from the data.