Make  Data  Count  -­‐  Demo  

MDC  Team:  PLOS,  CDL,  DataONE  Jennifer  Lin  &  Ma,  Jones  

Make  Data  Count  

Partners    California  Digital  Library,  PLOS,  and  DataONE  

NSF  Grant  Record:  Grant  No.  1448821  Proposal  PDF  in  eScholarship  repository  

Project  page  h,p://arCclemetrics.github.io/MDC/  

Prototype  h,p://dlm.plos.org    LagoGo  soHware  is  Open  Source  h,ps://github.com/arCclemetrics/lago,o     2  

Scholars  access,  share,  cite,  &  reuse  papers  in  many  ways.    ArCcle-­‐level  metrics:  

3  

figshare

PMC Europe PMC Europe Data- base Citations DataCite

Reddit

ScienceSeeker

F1000 Prime

Diversity  of  acCvity  on  papers  

But,  data…  •  Also  1st  class  scholarly  object  •  Broader  role  in  research  process  •  Has  its  own  use  &  reuse  profile  •  Infrastructure  services  to  collect  metrics  are  lagging  

•  Common  best  pracCces  are  not  enshrined  in  research  communiCes  

Mechanisms  for  mining  citaCons  

þ ArNcle    ☐ Data    ☐ SoHware  

Mechanisms  for  usage  collecCon  How  do  we  count  usage  stats?    

Independent  downloads?  EnCre  package?  •  Sum  downloads:  1733  •  Average  downloads:  346  •  Maximum  downloads:  586  •  Whole  package:  35  

How  do  we  count  across  versions?  •  Sum  downloads  across  all  version?  •  Average  downloads  across  versions?  

 

Only  some  objects  change  in  a  new  package  

How  do  we  standardize  usage  collecNon  pracNces  to  compare  across  providers?  

•  Standard  means  of  reporCng  usage  of  arCcles  •  COUNTER  reports  remove:  – Web  robots  from  search  engines  –  Repeat  visits  in  short  Cme  window  (double  clicks)  – All  accesses  from  Python,  Java,  curl,  wget,  etc.  

•  ScienCsts  frequently  use  these  to  access  data  •  COUNTER  issues  with  composite  objects  •  COUNTER  issues  with  versioning    

We  will  propose  changes  to  COUNTER  for  data  and  data  package  downloads.  

h,p://www.projectcounter.org/  

Lago,o  LagoGo  soHware  is  Open  Source  h,ps://github.com/arCclemetrics/lago,o    

Linked  Ref  Manager  Bookmarks  

Linked  data  citaCon  

Linked  Wikipedia  references  

Make  Data  Count  Project  Plan  1.  DLM  Field  Research      What:  Surveys,  interviews,  focus  groups  to  determine  requirements  for  DLM    Output:  metrics  design  &  requirements    

2.  Data  Usage  Tracking    What:  Extend  DataONE  usage  tracking  capacity    Output:  extended  usage  API  (COUNTER-­‐based)  

3.  Data  AcCvity  AggregaCon    What:  Formulate  a  set  of  metrics  to  text;  extend  technology    Output:  DLM  applicaCon    

4.  DLM  IntegraCon  &  PresentaCon    What:  Develop  tools  for  the  community  to  use  metrics    Output:  DLM  Reports  applicaCon  &  widgets  Bibliometric  Analysis    What:  Analyze,  write  up  results  from  project    Output:  final  report  &recommendaCons  

Tell  the  broader  story:  create  an  aggregate  narraCve  of  research  acCvity  by  bringing  metrics  on  data  &  

arCcles  together  Create  a  report    •  Keyword  •  Author  •  InsCtuConal  affiliaCon  •  PublicaCon  date  •  Subject  areas  •  Funder  

Stanford  University  +  NaNonal  Cancer  InsNtute  

Visualize  the  data  

Re-­‐imagine  the  problem  of:  •  Discovery  •  Reproducibility  

Thank  you  

Jennifer  Lin,  PLOS  jlin@plos.org  

Ma,  Jones,  DataONE  jones@nceas.ucsb.edu  

Make  Data  Count  Team  h,p://arCclemetrics.github.io/MDC/  

•  MarCn  Fenner,  PLOS  •  Kristen  Ratan,  PLOS  •  John  Chodacki,  PLOS  •  Dave  Vieglais,  DataONE  

•  Patricia  Cruse,  CDL  •  Carly  Strasser,  CDL  •  John  Kratz,  CDL  

Code  usage  &  raCngs