Technology Helps Job Search

Posted July 5th, 2012 at 2:20 pm (UTC-5)
Leave a comment

Friday, government experts are expected to report that the U.S. unemployment rate stayed around 8.2 percent, and the economy had a net gain of about 95,000 jobs.

Economists say this meager rate of job growth is not enough to reduce the jobless rate.

But experts say even at a time of high unemployment, some positions are going unfilled because companies cannot find people with the right skills to do particular high-technology jobs.

A company called Monster Worldwide is using some new technology it says can help solve that problem.

Company official Earl Rennison says computer programs called “semantic search” are able to scan vast amounts of data, including millions of resumes, to help match skilled people with vacant jobs.

Computer searches have been available for years, but he says they just matched certain key words, which Rennison says could give poor results if the word has multiple meanings.

Someone's name may be Ford, John Ford, or they may work for the Ford Motor Company, or go to, or have attended, the Ford School of Public Policy. So being able to differentiate these multiple different senses of a word is important for a semantic engine to understand.

He says semantic search is designed to pick the right match from the words and phrases around a key word. Rennison says this analysis of context helps it get the right match even if a job posting and resume use different words to express the same concept.

Some words have multiple ways of expressing them. For example, a programmer may be called a programmer or a software engineer or software developer or a web developer, etc. and there are multiple synonyms for the same word.

Rennison says the system was built in English, and has been translated into French, German and Dutch. He says his colleagues are working on a Chinese(Mandarin)version as well.

While semantic search is used by some other companies, Rennison says the one at Monster Worldwide is uniquely powerful and effective.