That's not the proper way to do it. Instead, you better first determine the purpose of the data that you collected in your system. Then, you'll figure out what functionality – or system behaviour – you'll need to fulfil these purposes with the new system, and which data is required.
Finally, you'll figure out how to compose these datasets with the data from your source system. So, while undertaking data migrations, try to avoid the temptation to start with data mapping. First think about what you'll need the data for and how you'll achieve these objectives. Only then will it be time to determine the data requirements and data mappings.
The first step is called 'functional mapping' and is characterised by looking from your existing system and it's data, towards the new system. The second step is the actual data mapping and is, in fact, looking backwards from the new system’s data requirements to the existing system and the data collected.