# Number of Scientific Articles per Country        
        
date of publication: 2015-03-28          
data source: Web of Science Tagged Data for SCI, SSCI, AHCI file creation 2015-01-09          
        
## data source:        
        
  last delivery of Thomson Reuters as "Web of Science Tagged Data" files      
  incrementally added to our base data starting with publishing year 1980      
  only recent versions of corrected records used      
        
### last delivery      

```        
  |file name   |creation|prod week        |product
  |------------|--------|-----------------|----------------------------------------------------------------------
  |IF3N1503    |20150109|IP201503-IP201503|Science Citation Index Expanded , SciSearch
  |IF3N1503    |20150109|IP201503-IP201503|Arts & Humanities Citation Index, Arts & Humanities Search
  |IF3N1503    |20150109|IP201503-IP201503|Social Sciences Citation Index , Social SciSearch (see ‘J’ before 1995)
  |IG3N1503.GAP|20150109|IP201503-IP201503|Science Citation Index Expanded , SciSearch
  |IG3N1503.GAP|20150109|IP201503-IP201503|Arts & Humanities Citation Index, Arts & Humanities Search
  |IG3N1503.GAP|20150109|IP201503-IP201503|Social Sciences Citation Index , Social SciSearch (see ‘J’ before 1995)
  |IC3N1503.COR|20150109|IP201503-IP201503|Science Citation Index Expanded , SciSearch
  |IC3N1503.COR|20150109|IP201503-IP201503|Arts & Humanities Citation Index, Arts & Humanities Search|
  |IC3N1503.COR|20150109|IP201503-IP201503|Social Sciences Citation Index , Social SciSearch (see ‘J’ before 1995)
```

  Copyright (c) 2015 Institute for Scientific Information      
        
        
### WoS tagged data        
RP  corresponding affiliation  WoS Reprint Address (Sub-record of Source Item Record. Not required, max occurrence 1)    
NU  country  subfield NU = country    
        
### raw data processing        
country normalization    WoS country entries (original field NU, currently 256 unique variants) were semi-manually normalized    
    to ISO 3166-1 alpha-2 codes (212 unique variants)    

## basic sql statement 

```
select woad.nu, wois.py, 
       count(distinct woad.fk_source_item_records) nrec_tot,
       count(distinct case when woad.type = 'RP' then woad.fk_source_item_records else null end) nrec_rp
from wos_research_reprint_addresses woad
join wos_source_item_records woit
 on fk_source_item_records = pk_source_item_records
join wos_source_issue_records wois
 on fk_source_issue_records = pk_source_issue_records
where dt in ('@ Article', 'R Review')
group by woad.nu, wois.py;
```
 
note: the sql given summarizes the direct logic based on a pure relational transformation of the tagged data  
data actually were queried after various in-house transformation and normalization steps
    
## data files

### table "01_global"

```
subset        
  01 articles total                     ... total number of Web of Science items with document types (dt) "article" and "review"     
  02 articles with affiliation entries  ... subset of 01 with affiliations attached to the article    
  03 sum of articles per country        ... total number of articles split by affiliation country (multiple countries per article) 
                                            based on 02    
  04 corresponding affiliation articles ... subset of 02 with articles that have a corresponding affiliation marked

breakdown by publishing year        
  a2004 ... a2013                       ... numbers per year the articles where published    
```

#### table "02_countries"
```
code                                    ... ISO 3166-1 alpha-2 code for country    
country                                 ... name of country    

breakdown by publishing year        
     t2004 … t2013                     ... total number of articles and reviews in Web of Science    
     c2004 … c2013                     ... number of articles and reviews with corresponding affiliation in given country    
     s2004 … s2013                     ... share of corresponding author articles [%]    
```

## stastistical analysis

```
sesam_readcsv_case('md', 'mpdl_rio_wosartbyco_country.tsv');

# increase of total number of articles [%]

summary(md$t2013[md$t2004>0]/md$t2004[md$t2004>0]*100-100);
    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
 -100.00    68.22   120.10   355.40   200.00 33040.00 
summary(md$t2013[md$t2009>0]/md$t2009[md$t2009>0]*100-100);
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
 -83.33   18.39   39.04   61.09   67.63 1400.00 

# share of corresp. to total articles [%]

summary(md$s2004);
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
   0.00   31.00   55.50   50.97   72.00  100.00      16 
summary(md$s2009);
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
   0.00   30.00   51.50   49.89   69.00  100.00      14 
summary(md$s2013);
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
   0.00   25.00   41.00   42.43   60.50  100.00       5 

# third degree polynomial

mdx = md[md$t2013 > 10,]; # for countries with a minimum of 10 articles in 2013 

lm(formula = mdx$s2013 ~ poly(log10(mdx$t2013), 3))

Coefficients:
               (Intercept)  poly(log10(mdx$t2013), 3)1  poly(log10(mdx$t2013), 3)2  poly(log10(mdx$t2013), 3)3  
                     45.54                      242.92                      -19.02                      -21.81 
```