The agency said any move must reflect the commercial and privacy concerns of individual food business operators and FSA staff.
Comments are invited before 17 March on the proposals which also identify publication timelines.
The proposal for milk and dairy inspection scheduling data is retrospective publication after inspection of the premises.
For the list of registered milk production holdings publication would be monthly.
Enforcement data would be published retrospectively after the appeal process has passed at quarterly intervals.
A proposed date of the end of March is subject to acceptable levels of aggregation and anonymity.
Contractor key performance indicators would be published quarterly after the contract notice period has expired.
In April 2016 the FSA took on board a government move to make more of data, into a target to publish 95% of all its datasets by 31 March 2017.
At the end of quarter three 2016/17, FSA had published 92 of 269 datasets (34%).
The FSA announced last year that it would fund a research fellow in Data Science and Food at University College London (UCL).
The fellowship will focus on developing analytics methods for tracking and characterising foodborne illnesses.
Professor Guy Poppy, chief scientific adviser at the FSA, said at the time that data science has the potential to transform what it does.
“Working with UCL’s Big Data Institute, we really are putting consumers first in all that we do, to ensure food is safe and what it says it is,” he said.
“We all eat at least a few times every day, so food is a topic well-suited to testing the effectiveness of data science approaches. It is an important collaboration for all parts of the FSA business.”
Transparency vs commercial interests
The aim is to have this list of data sets available with a click through to live data which is open.
For example, the user will get to the Food Hygiene Ratings Scheme (FHRS) data page through the FHRS data entry in the list.
“Within FSA Field Operations and Operations Assurance, which covers all audit, inspection and enforcement activity in slaughterhouses and cutting plants; and dairy hygiene and wine inspection schemes we have identified 20 individual datasets,” said the agency.
“In developing these open data proposals in relation to FSA Field Operations and Operations Assurance Division datasets we believe we have struck the correct balance between transparency and protection of commercial interests both within UK markets and in competition for markets abroad.
“We have adopted levels of data aggregation, which also addresses welfare concerns of FBO and FSA staff. Individual plants and staff members will not be identifiable from the release of this data beyond what is already in the public domain.”