RESUME PARSER FOR CYRILLIC ALPHABET
Problem
One of the biggest HR companies in Eastern Europe came out with the problem of not having enough resources to document and extract information from all of the resumes they receive hourly. Also, the percentage of errors as a result of the manual work was quite large. The biggest challenge was to create a CV parser that will be able to parse and store information in the Cyrillic alphabet.
Solution
Creating rule-based and machine learning model for extracting the following information from every CV:
- Name and last name of the applicant
- Date of birth and address
- Education
- Work experience
- Interests
Results
The parser was tested on 20.000 resumes and provided results with an average of 78% accuracy for each type of information.
