The service was delivered according to the following phases:

  1. Preparation – cleaning the dataset, extraction of relevant information from consultants’ CVs, customers’ feedback and project’ description data.
  2. Extraction of features for both consultants and projects in terms of numerical Representation. Standard Natural Language Processing and Text Analysis techniques were used for summarization
    1. Textual summary of each consultant’s CV along with his/her skills extracted
    2. Textual summary of each project along with required skills extracted

Here is detailed list of extractions:

  • Demographic data – Place of birth, Place of living, Country and language
  • Education history – Institution, Degree, Period of education, Build Faculties Database and Ranking
  • Work experience – Company, Period, Years of experience, Build companies database and Ranking,
    Linking hard skills and soft skills to work experience, Introducing new entities –
    Position – for instance, someone is software developer, product manager, testing
    engineer and similar.
  • Work Style – Soft skills extraction, Building similar solution as for the hard skills.
  • Final submission – System tuning, Integration

The deliverables of each phase contained:

  • API for retrieving phase-specific information
  • GUI demo for viewing phase-specific information
  • Evaluation of the extraction algorithms