Fill in an GitHub issue with the paper and some associated tags you think are right for it, as in here.
Suppose you want to add the paper called "Practical Covertly Secure MPC for Dishonest Majority - Or: Breaking the SPDZ Limits"
As the current pipeline stands, if you want to add a paper you need to follow the next steps:
- Find its semantic scholar id (s2id), for example: https://www.semanticscholar.org/paper/Practical-Covertly-Secure-MPC-for-Dishonest-Or%3A-the-Damg%C3%A5rd-Keller/e69e44f837662fdc5c49a7ba40666ba7bd5980c0
The s2id of a paper consist in the entire digest before the end of the link, i.e.: e69e44f837662fdc5c49a7ba40666ba7bd5980c0
- Look for the json information corresponding to the s2id in
crypto-semantics-data.json, eg:
...
{
"label": "DKL+12",
"abstract": "SPDZ (pronounced “Speedz”) is the nickname of the MPC protocol of Damgard et al. from Crypto 2012. In this paper we both resolve a number of open problems with SPDZ; and present several theoretical and practical improvements to the protocol. In detail, we start by designing and implementing a covertly secure key generation protocol for obtaining a BGV public key and a shared associated secret key. We then construct both a covertly and actively secure preprocessing phase, both of which compare favourably with previous work in terms of efficiency and provable security.",
"id": "e69e44f837662fdc5c49a7ba40666ba7bd5980c0",
"venue": "ESORICS",
"s2Url": "https://semanticscholar.org/paper/e69e44f837662fdc5c49a7ba40666ba7bd5980c0",
"authors": [
"Ivan Damgård",
"Marcel Keller",
"Enrique Larraia",
"Valerio Pastro",
"Peter Scholl",
"Nigel P. Smart"
],
"title": "Practical Covertly Secure MPC for Dishonest Majority - Or: Breaking the SPDZ Limits",
"year": 2012
},
...Check if the s2id is in tagged_output.json. If it is and you actually want to delete the entry
from the Kumu map then set the flag isMPC: 0.
If the s2id is not in tagged_output.json then add this entry at the end of of tagged_output.json.
with the corresponding isMPC flag:
{
"label": "DKL+12",
"abstract": "SPDZ (pronounced “Speedz”) is the nickname of the MPC protocol of Damgard et al. from Crypto 2012. In this paper we both resolve a number of open problems with SPDZ; and present several theoretical and practical improvements to the protocol. In detail, we start by designing and implementing a covertly secure key generation protocol for obtaining a BGV public key and a shared associated secret key. We then construct both a covertly and actively secure preprocessing phase, both of which compare favourably with previous work in terms of efficiency and provable security.",
"id": "e69e44f837662fdc5c49a7ba40666ba7bd5980c0",
"venue": "ESORICS",
"s2Url": "https://semanticscholar.org/paper/e69e44f837662fdc5c49a7ba40666ba7bd5980c0",
"authors": [
"Ivan Damgård",
"Marcel Keller",
"Enrique Larraia",
"Valerio Pastro",
"Peter Scholl",
"Nigel P. Smart"
],
"title": "Practical Covertly Secure MPC for Dishonest Majority - Or: Breaking the SPDZ Limits",
"year": 2012,
"isMPC": 1
}Once this is done please submit a pull request and I will make sure to re-run a script that adds automatically the connections in the visual map.
To get a list of COED papers take the following steps:
- Fetch semanticscholar local database of cryptography papers and output the result into a json by running:
python kumucollector.py --output 0
This will output a big file as `kumu_output.json'
- Train a model using the some partially labelled data by running
python trainer.py
- After having the ML model now run obtain the new COED paper classification using
python predictor.py
This should output a json called ``predicted_output.json`.
- If you want to restrict the graph to, say, 128 citations execute:
python kumucollector.py --output 2 --limit 128
The last command will use predicted_output.json together with the connections
from kumu_output and will output the kumu graph in outKumu128.json.
In order to avoid losing the tag information we recommend running
python tag-adder.py --in crypto-semantics-data.json --out predicted_output.json
after a re-run of python kumucollector.py --output 0
Perhaps there is some wrong information about a paper such as author names, conference name,
year appeareance, etc. Add it to corrections.json with the fields you wish to correct
along with its semantic scholar id. Then re-run python kumucollector.py --output 0
Question: Why is not paper X in there?
Answer: Most likely because it's indexed wrong by semantic scholar. The way we built this database was by selecting for papers which appeared in CRYPTO, EUROCRYPT, CCS, etc. Sometimes semantic scholar does parses the venues incorrectly, or even author names, sometimes it has duplications of the same paper so the number of citations might be slightly different than what you see on Google Scholar.
Q: Where can I see this marvellous map?
A: You can find it on Kumu here.
Q: Are there any other cryptography maps?
A: Yes, a very more detailed map on FHE literature made by Ilia Iliashenko. Check it out here.
We would like to thank Peter Sebastian Nordholt for posting the scripts used in https://guutboy.github.io/ on GitHub. We would also like to thank the COSIC group for their helpful feedback on initial version of the map.