In Paradigm Shift in Job Search I claimed “…job search will never be the same.” Finally, we can accurately assign precise structured occupational codes to jobs after they’ve been indexed from multiple sources. Equally important, we can interpret search strings using the same taxonomy and technology … placing each job search in context.
Here’s an example:
Let’s say a jobseeker is searching for a ‘computer network support’ position using a leading job board. This job board accepts the user’s search string, parses the string into keywords, and then matches the words and phrases against all jobs in their index.
If you try this on your favorite job board you’ll likely find that some jobs returned by the search are an excellent fit; but you’ll also find many jobs, containing your keywords, are totally out of context for a ‘computer network support’ position.
Not surprising really. Most of today’s job search engines use some variant of the ‘bag of words’ approach to search. Problem is, once the words are in ‘the bag’, context is lost.
The alternative – the first part of our paradigm shift in job search – adds context to each search string.
The image at right (click the image for a larger view) shows the best-fit occupation codes selected by our O*NET-SOC AutoCoder software for the ‘computer network support’ search string. Notice that these occupation codes are ranked by relevance score.
Now, instead of searching all jobs in our index for the words ‘computer network support’, we can search for all orders classified within the five occupational categories deemed most appropriate for the search text. In addition, we can cluster the search results by the above five categories.
Adding context to search isn’t a new concept – what is new is the ability to automate the process at a high degree of accuracy. For more information about autocoding jobs and search strings, see my O*NET-SOC AutoCoder blog.
Comments