SWEPUB - ORCID¶

version 0.8

This notebook
SWEPUB
- Kundo question were they recommend download the ZIP file to access data in SWEPUB --> JSON 10.81 Gbyte
- datamodell/swepub-bibframe
- Twitter SwePub
- SPARQL SWEPUB feels SPARQL is better than download a zipfile ?!?!?
  - Finding nr records per schools
  - authorityRecords/organizations.sparql --> query
  - publicationChannels --> query
  - Finding persons with ORCID ???

see at the end where we find an ORCID in the JSON file looks like not everyone has an ORCID...

Magnus C Persson 0000-0003-1062-2789 he is same as WD Q88134673 --> Scholia ->
- co-authors
- graf
Wikidata
- ORCID property P496 on 1 546 332 objects
  - 0000-0002-5494-8126 --> WD query haswbstatement:"P356=0000-0002-5494-8126"
- DOI property P356 on 25 609 492 objects
  - doi/10.1186/S13321-016-0161-3 --> WD query haswbstatement:"P356=10.1186/S13321-016-0161-3"
- query using SPARQL
- Scholia - tool for citation graphs of data in Wikidata
  - DOI link doi/10.1186/S13321-016-0161-3
  - ORCID link [orcid/0000-0002-5494-8126](https://scholia.toolforge.org/orcid/0000-0002-5494-8126]

In [ ]:

# Try to get ORCID from SWEPUB
# see https://kundo.se/org/swepub/d/api-for-amnesklassificering/#c3571837
import pandas as pd
import json
import time
start_time = time.time()
filename ="data/swepub-duplicated-2020-07-05.jsonl"
filestore ="data/swepub-duplicated-2020-07-05_1.pd"

print(time.ctime())
df_chunk = pd.read_json(filename, lines=True, chunksize=10000)
chunk_list = []
for i, chunk in enumerate(df_chunk):
    chunk_list.append(chunk)
print("--- %s seconds ---" % (time.time() - start_time))
# concat the list into dataframe
df_concat = pd.concat(chunk_list)
print("--- %s seconds ---" % (time.time() - start_time))
df_concat.info()
#df_concat.to_pickle(filestore)
#print("--- %s seconds ---" % (time.time() - start_time))

Thu Jul 16 16:56:56 2020

In [ ]:

pd.set_option("display.max.columns", None)   
df_concat["instanceOf"]

In [ ]:

pd.DataFrame(df_concat["instanceOf"].tolist()) 

Try to find ORCID and DOI¶

In [ ]:

instanceOfdf = pd.DataFrame(df_concat["instanceOf"].tolist()) 

In [ ]:

instanceOfdf

In [ ]:

pd.DataFrame(instanceOfdf["genreForm"].tolist()) 

In [ ]:

instanceOfdf.info()

In [ ]:

pd.DataFrame(instanceOfdf["hasTitle"][1:10].tolist()) 

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][1:10].tolist())

In [ ]:

pd.DataFrame(instanceOfdf["hasNote"][1:10].tolist())

In [ ]:

pd.options.display.width = 0
pd.DataFrame(instanceOfdf["contribution"][1:10].tolist())

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][1:10].tolist()[0])

we have 35 authors but looks like no ORCID¶

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][1:10].tolist()[0])["agent"]

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][1:10].tolist()[0])["agent"].tolist()

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][1:10].tolist()[1])["agent"].tolist()

In [ ]:

pd.options.display.width = 0  
pd.DataFrame(instanceOfdf["hasTitle"][1:10].tolist()[2]) 

In [ ]:

pd.DataFrame(instanceOfdf["hasTitle"][1:10].tolist()[2])

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][1:10].tolist()[0])["role"].tolist()

In [ ]:

pd.DataFrame(instanceOfdf["electronicLocator"][1:10]) 

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][1000:1010].tolist()[1])["agent"].tolist()

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][2000:2010].tolist()[1])["agent"].tolist()

Looks we have an ORCID¶

Magnus C Persson 0000-0003-1062-2789 he is same as WD Q88134673 --> Scholia ->
- co-authors
- graf

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][2000:2010].tolist()[1])["agent"].tolist()[0]["identifiedBy"]

In [ ]:

pd.DataFrame(instanceOfdf["contribution"][2000:2010].tolist()[1])["agent"].tolist()[0]["identifiedBy"][0]["value"]

In [ ]: