Develop a resume & job scoring and matching strategy

The goal is to match the best resumes to a job opening. Or reverse, match jobs in the database to a resume. Resumes have words that are associated with job titles, skills, technologies, certifications…etc. These words or “attributes” can be counted for occurrence or placement in a resume (is higher in the resume better?). Further, these attributes can be hierarchical in nature. Example: Java is a development language. Struts and Hibernate are Java frame works. These could sit under “Java” in the hierarchy. By having more of the “children” attributes, the resume could be quantified as more qualified at Java. The job openings could also be scored similarly for attributes. In addition, there could be many hierarchies, covering industries, technologies and job titles. Question 1: What is the most effective way of “scoring” a resume? Is it to use attributes? Should the score reflect multiple attributes as opposed to an overall score? Is there an alternative way which is more effective? Question 2: What is the pseudocode necessary to match a multi-attribute scored resume to job openings? Deliverable: Develop a scoring & matching methodology and pseudocode suitable for a developer to use