Problem Statement ID :- 1652
Problem Statement Title :-- Extraction and Verification of Information from semi-categorised data.

• Background: Recruitment and Assessment Centre (RAC) under DRDO, Ministry of Defence invites online biodata applications with requisite eligibility supporting documents from candidates. Verification of these documents is presently a manual process. • Description: The verification of submitted documents such as educational certificates, marksheets, GATE scorecard, experience, caste, EWS, PwD certificate, etc. by applicants, as applicable against advertised vacancies, is presently carried out manually. These documents, generally available in image or PDF form, may be in languages other than Hindi or English. In order to verify the information filled by applicants in their biodata applications can be verified using Intelligent Document Processing (IDP) techniques involving Machine Learning Deep Learning, Artificial Intelligence, Natural Language Processing, etc. • Expected Solution: The extraction of information from submitted documents can take place at the time of application form submission and alerting applicant in case of any mismatch. Thereafter, while screening of applications more robust data extraction techniques may be employed to further avoid any chances of error. The extracted information may be presented using modern business intelligence tools thus highlighting the efficiency of process. The solution should be able to process documents with Three Sigma accuracy with a speed of not more than 3 seconds per document.

#smartindiahackathon #sih2024 #sih #smartindiahackathon2024 #thoughtsrevoked

свернуть