I see this price is not in EDDB so this will get cleared out tonight. I do have some filtering but it probably should be more aggressive. Ideally we could make a curated channel within EDDN that filters out the bogus price data from the start, so each tools doesn't have to implement its own filtering.
A curated channel has been discussed way back in January or something. Because of the K.I.S.S principle of the EDDN it was dismissed. It was decided that curating data should be done in the clients/receivers. The eddb (which dumps you use) is heavily curating data btw as is Maddavo. You could write them for info how they did it. Maddavo also had a thread for it:
Price sharing - price sanity check algorithms
However recently EDDN got an update to v0.4 (done by Anthornet) where the can be more relays. If I've interpreted it correctly all data is send to the main gateway and then distributed to one or more relays. One such a relay could be publishing ship information or be a curated data channel. But then EDDN should be extended with extra logic or someone should be sending curated data to the EDDN. An alternative: EDDN receives data but before giving it to a relay it sends data to a cleaning service which curates the data and then sends it back to the EDDN. That data is then send to the relay(s). Not practical with schema V1 but with schema V2 (one msg holds all prices for a complete station) it could be do able. Anthornet is currently on vacation so don't know what is exactly possible. Not sure if curating will end up in the EDDN or that we should even want that. K.I.S.S and such. It might be worthwhile to restart the discussion. Feel free to create a post in the EDDN thread.
A pragmatic approach. In an EDDN message one can see what software did send the data. You could filter on that. As in decide that the
OCR Error Free tools aka
Mobile API based tools data is clean (its comes directly from the web-api of FD) and use that. Discarding the rest of the messages. But then you will lose data. If you look at the
EDDN-monitor you notice that OCR data is in the minority lately and that far more data (factor 4-5) is coming from the OEF tools. Atm the most is coming from
EDMC which btw supports all public trading tools.
TLDR; Given the fact the EDDN v0.4 has been released and the support for it for schema V2 it is worthwhile to restart a discussion about curating data in the EDDN. In the past we decided against it.
Note: You might consider putting the OEF tools on the OP of ETN. That way it gets more exposure, more users and ETN will receive more pristine and 100% correct data. Please read about the status and legality of the OEF tools
here.