Text Mining in E-Recruitment

Online recruitment is a huge industry. The cost of recruiting online is a fraction that to hire employees through the traditional methods (such advertising in magazines or news papers and registering job vacancies with employment agencies). The older, slower, paper-driven methods of recruitment have been transformed by use of the web. Now much of the recruitment process prior to a face-to-face interview is performed using email or through a website.

Most companies advertise their vacancies on the web. Some companies use a well-known website that specializes in matching candidates and requirements. Other companies publish a list of vacancies independently. Depending on the market and the demand, a company may be flooded with resumes. Dealing with hundreds of resume is laborious, and in some cases, thee law does not allow employers to delete e-mail responses and each applicant’s resume must be filled. Sending resumes electronically is easier than mailing a paper resume and there is no reason to believe that the trend of sending electronic resumes over paper resumes will be reserved. If employers had tools to quickly sift through a pile of resumes, the recruitment process could be faster.

In the preliminary manual scan of resume, a recruiter looks for typos, educational qualifications, buzzwords, employment history, frequency of job changes, job titles, and other personal information. Automatically extracting this information can be the first step in filtering resumes. An employer may have a minimum requirement for educational qualifications and work experience. Based on the extracted attributes, resume can be grouped into categories. A higher weight may be assigned to resumes received by referral. Resumes with many common buzzwords as well as approximately similar numbers of years of work experience and educational qualifications may fall into one category. Likewise, other combinations of attributes can form other categories. Some attributes, such as years of experience or particular buzzwords, may be assigned higher weights than educational qualifications, for instance.

An ideal resume of a candidate for vacancy with the perfect educational qualifications and work experience can be constructed. Attributes of submitted resumes can be compared with the attributes of the ideal resume to generate a score. Finally all submitted resumes could be ranked based on this score. These and similar methods can reduce the burden of managing resumes.

Although its difficult for employers to handle large volume of resumes, it is also difficult for applicants to find appropriate vacancies. Often vacancies announced on a company’s web site are not announced through a large broker specialized in a placement. Such vacancies may not be noticed by prospective applicants. Periodically crawling web pages that announce vacancies and automatically matching the findings with personal requirements is one way of monitoring recruitment at a company.

Vacancy listings are described in a somewhat standard format. A job vacancy template can be filled with extracted information such as the title, location, salary, experience, and educational requirements from a job listing web page. A candidate can target vacancies where there is a good match between requirements and his qualifications. Several newsgroups on the web specialize in such announcements.
A deeper search by a candidate of the company’s web site may reveal opportunities that are not listed in the vacancy pages. Such hidden information is hard to uncover without reading many web pages. Information on products, tools, and people involved in projects of interest provides leads in the hidden job market. Automating this collection process can speed up a job search and improve a candidate’s prospects.

Artificial Intelligence, Machine Learning, Natural Language Processing and Text MiningTechniques are delivering good solutions in this industry. Most of the recruitment solutions are using the above techniques to capture the vital information about the candidates and job postings. Automatic Job Searching and Matching is one of the hot field to make money.