Machine learning and Natural language processing for Automating Systematic Review Development and Meta-analysis

This Challenge is issued under the BCIP Challenge Call for Proposals 002 (EN578-DB1800/A).

Please refer to the Solicitation Documents.

Challenge Name: Machine learning and Natural language processing for Automating Systematic Review Development and Meta-analysis

Challenge Notice: EN578-DB1801

Priority Area: Military Component

Maximum Contract Value: $675,000 (taxes extra, if applicable)

Challenge Sponsor: Public Health Agency of Canada (PHAC) / Department of National Defence (DND)

Funding:

This Challenge is being fully funded by PHAC.

PHASE I – Fiscal Year 2018/2019: The available funding for Phase I is approximately $225,000 (taxes extra, if applicable) per contract. Up to two contracts could be awarded.

PHASE II – Fiscal Year 2019/2020: The available funding for Phase II is approximately $450,000 (taxes extra, if applicable). Phase II may be exercised for one of the contractors initially awarded a contract under Phase I.

All funding is in Canadian Dollars. This disclosure is made in good faith and does not commit Canada to contracts for the approximate funding stated above.

Background/Summary:

As the Government of Canada's lead federal agency for disease prevention and management, the Public Health Agency of Canada (PHAC) is mandated to promote and protect the health of Canadians. In order to support health care professionals, policy makers and citizens in making informed decisions, PHAC must conduct timely, accurate and comprehensive analysis of evidence. To pursue its mandate, PHAC requires an increased capacity to conduct rapid (real-time) and comprehensive evidence reviews.

A machine learning (ML)/natural language processing (NLP) platform is required to support the systematic reviews of accumulating evidence related to clinical, public health policy and program decision making.

Within this context, an ML/NLP Innovation is sought to support the undertaking of systematic, structured, and transparent searching, selection, review and grading of relevant peer-reviewed published literature based on a pre-defined research question. The Innovation must have an ability to extract and export key results of published studies to pre-defined data/evidence tables.