Going down the rabbit hole a little further...
I'm going to guess the value "netflow.network.direction" you are hoping for is "flowDirection" from
the original IPFIX RFC
5.11.6. flowDirection
Description:
The direction of the Flow observed at the Observation Point.
There are only two values defined.
0x00: ingress flow
0x01: egress flow
Abstract Data Type: unsigned8
Data Type Semantics: identifier
ElementId: 61
Status: current
Sounds interesting, right? Who wouldn't want to know if their flow is in or out?
Well, the problem is that this is a property of a unidirectional flow.
Let's take the example of a simple http connection between a Host1 and a Host2. This will create 2 unidirectional flows, one from Host1 to Host2 egress and one from Host2 to Host1 ingress. What you probably want to know is which host is the Server and which is the Client so that you can assign a bidirectional flow direction. To do this you need to apply more data than you have here. You could naively assume that any flow "egress" from your WAN port is towards a server and initiated by the client (if you have effective and restrictive firewall filters, I hope), or you could guess that any destination port < 1024 is toward the server etc, etc.
The basic point is:
1) IPFIX sends you info on unidirectional flow information
2) It turns out the "direction" of a unidirectional flow is not very interesting
2) A lot of information we would like depends on creating bidirectional flow info (top talkers etc)
3) Unidirectional flow "direction" doesn't help much to create bidirectional flow Client - Server "direction"
4) Creating bidirectional flows is not trivial and usually uses a lot of data outside of IPFIX, like assumptions about your network setup, and uses a bunch of code on whatever collector you use.
Also, you can derive flow "direction" from other IPFIX data
P.S. if youre interested in interface stats or direction through the router they are usually based on INPUT_SNMP and OUTPUT_SNMP