Personalized Job Search Experience!

Personalized Job Search Experience!

Job portals continue to remain the most widely used recruitment mechanism.

The question emerging in the changing job scenario however is their effectiveness in finding the right job for candidates.

Job portals were the first generation products of the Internet when the internet was all about collecting information and creating access.

The idea was as the internet reach expanded, more and more people and opportunities could get onto this platform and find each other.

But, before we debate the relevance of job portals, it is important to understand the value job portals were creating.

Overtime the problem in helping people get jobs or helping corporates find people has changed from ‘simply finding the people’ to ‘locating the right people’.

The utility of the internet today is no longer restricted to access to important information, but in finding the right information.

Image for post

In this scenario, searching through Traditional Job Portals sees two major hurdles:

  • Lack of useful information.
  • No recommendations based on Resume.

The problem of too much information:

Unfortunately, information available through job titles is not enough to do an intelligent match.

Job seekers/applicants see very similar jobs of hundreds of descriptions, making it impossible to figure out which job seems to be relevant or better.

If we see this situation from a job seekers’ perspective, a job portal has hundreds of open jobs.

For a candidate with basic entry-level skills it becomes fairly hard to analyze which jobs is he or she ready for and should apply to.

Eventually, they apply to too many jobs which makes the overall process of job hunting inefficient.

Also, most organizations use automated mechanisms to reach out to these candidates and get them to go through a process.

Again a phenomenal waste of time for them to go through multiple processes with really no useful information/feedback.

In this process, they might also miss out on the right jobs for them.

Any marketplace works effectively when there is value for both sides of the ecosystem, in this case — Job Seekers or Employers.

As the number of people and opportunities on these platforms is increasing, effective matching is breaking.

With the proliferation of the internet and the rise in the number of candidates and jobs on the platforms, their effectiveness is in question and the platforms need to deploy smarter ways of Job Matching.

e.g. Java Developer may sound specific. But there are different types of Java developer:

  • Proficient in different Java systems (standard, enterprise, and mobile).
  • Some Java developers advance to lead or architect positions.
  • A senior Java developer may analyze complex problems, develop documentation, review coding, and evaluate the development process.
  • An architect, meanwhile, directs the project at the front end.
  • EE architect is among the highest positions a developer can attain.

Why are Traditional Job Search Portals not providing Personalized Matching

It takes a lot of research and effort to develop a system for personalized job matching. which includes:

  • A lot of historical and recent data on job postings.
  • Develop a model using the data, using some advanced NLP (Natural Language Processing).
  • The complex data structure of Resume. (Parsing resume is difficult as it does not follow definitive structure)
  • Transitional roles in Resume.

    Transitional roles mean: Career transition in your professional background and experiences. (During developing personalized Job Search Portal, one has to take into account Transitional roles, which make it difficult, as it takes effort in identifying the most recent Job Profile of Jobseeker.)

    • The use of NLP (Natural Language Processing) and A.I. takes a lot of effort.

    How is different?

    • Personalized matching for your skill-set and role.
    • Based on state-of-the-art NLP (Natural Language Processing) systemto parse job and resume.
    • Uses Entity Recognition Model to extract skills and other important features from both, Jobs and Resume.

    [Named-entity recognition (NER) is a subtask of information extraction that seeks to locate and classify named entity mentioned in unstructured text into pre-defined categories such as person, organizations, locations, time expressions, quantities, monetary values (currency), percentages, etc.]

    • Uses syntactic and semantic matching to recommend the best jobs.

    How built this model:

    • Beyond a simple keyword match, we used the cutting edge Entity Recognition Model.
    • We used a crawling mechanism to collect both historical data and recent data of job postings for Tech and Cybersecurity.
    • The data was then annotated by domain experts.

    This pipeline helped us in building a state-of-the-art Matching System.

    Personalized Job Matching by tries to find an automated answer.

    From this data, we develop a model that receives resumes from the candidates and requirements from the companies as inputs, builds up index using some advanced natural language processing, and retrieves the related jobs.

    Leave a Reply