Mismatch between author text fiels and paper text files for paper IDs?

[man...@aganitha.ai]

Hi everyone, a quick question about the database. While trying to map the authors with the papers in MAG database, I noticed that the paper IDs are mentioned in completely different format. For example, two random authors from mag_authors_0/mag_authors_0.txt are displayed as:

{"id": "2170957184", "name": "Keim Melville", "normalized_name": "keim melville", "pubs": [{"i": "974455535", "r": 0}], "n_pubs": 1, "n_citation": 1}
{"id": "2170957195", "name": "勇二 林", "normalized_name": "勇二 林", "pubs": [{"i": "569194110", "r": 0}], "n_pubs": 1, "n_citation": 0, "tags": [{"t": "Scheduling"}]}

Notice that the IDs of publications are 9 digit integers, e.g. 974455535. Now two of the papers from mag_papers_0/mag_papers_0.txt are displayed as:

{"title": "Modified K- Nearest Neighbor Classifier Using Group Prototypes and its Application to Fault Diagnosis.", "id": "0184e871-c4ea-4f21-a609-a67b5da32daf", ..........}
{"title": "Effect of MgSO_4 on the Expression of Fos Protein in the Dentate Gyrus of Hippocampus of Rats with Irradiation-Induced Brain Injuries", "id": "0184e87d-daf4-439c-94cf-76ec0528d4f1", ......}

Again notice the IDs of papers are actually 36 bit alpha-numeric values. My question is, how do I map the two datasets with those IDs? Is there some relationship between those IDs somehow?

Thank you

M.

[Fanjin Zhang]

Hi Manish,

Did you use the MAG papers of OAG v1? MAG papers in OAG v1 used long 36-bit paper id, but MAG papers in OAG v2 should use numerical id. This is because the old version of MAG published the 36-bit alpha-numeric paper id. We don't have the ID mapping relations of those two kinds of IDs, but perhaps you could try MAG papers of OAG v2, which should be consistent with the MAG author data of OAG v2.

[Manish]

Hi Fanjin, thank you for your reply. You may be right, I had a feeling that the authors data I had was from from v2 and papers from v1. Since now you have confirmed it, I guess I'll try to downlaod the right version. Thanks again.

Manish