Positioning Ukraine scientific excellence along bibliometric indicators

 

The following bilbiometric indicators to position Ukraine in terms of scientific excellence are discussed briefly as follows:

 

· the number of scientific Ukrainian publications among the 10% most cited publications worldwide as % of the total scientific publications of the country

 

· public-private Ukrainian co-publications by million population

 

· international scientific co-publications with authors from Ukraine per million population

 

Based on CWTS findings for Ukraine¹¹⁶, the quality of a country’s research basis can be approximated by the number of scientific publications among the 10% most cited publications worldwide as % of the total scientific publications of the country. According to this indicator, Ukraine shows regularly low shares in the available time series from 2002 to 2013, ranging around 3% (2002: 3.4%; 2013: 3.1%), with a negative trend from 2008 (3.1%) to 2012 (2.2%). The EU average meanders between 9.8% in 2002 and 10.5% in 2012, depicting a positive upwards trend. The Ukrainian performance lies below all EU cohesion countries (e.g. Bulgaria 2013: 3.5%; Czech Republic: 7.3%; Hungary: 6.5%; Poland: 5.0%; Romania 4.7%; Slovakia 5.5%) but also below Turkey (2013: 4.8%) and Russia (3.3%). Among all countries analysed for the IUS/EIS, it only surpasses the performance level of Albania. It goes without saying, that the gap between Ukraine and the best performing countries in this respect (Switzerland: 15.7% in 2013; Netherlands: 14.5%; USA: 14.0%; UK: 14.2%; Denmark: 13.3%) is even increasing (except USA).

 

Another important indicator used for assessing fundamental science-industry relations refers to the public-private co-publications by million populations. It indicates the level of knowledge-basedcooperation between academic and business R&D of a given country. The higher the indicator, the higher is this sort of knowledge-based inter-sector cooperation. While the EU average between 2008 and 2014 fluctuates quite heavily between 34.1 in 2008 and 33.9 in 2014 with a peak as high as 41.6 in 2011 and a low 33.9 in 2014, the Ukrainian time series data are stable, but at a very low level. It shows 1.0 in 2014 with a peak of 1.5 in 2010 and a low of 0.9 in 2013. Among the countries covered by the IUS/EIS similar low recent levels can only be found in Latvia, Turkey, the Former Yugoslav Republic of Macedonia, Montenegro, Bosnia and Herzegovina, Albania, Moldova and – India (which hints to an influence of size effects). The countries with the highest number of public-private co-publications by million populations in 2014 are Switzerland (217.6),¹¹⁷ Iceland (187.3), Denmark (143.5), and Sweden (107.8). Russia has a

 

 

116 Data from CWTS were provided by the European Commission.

 

117 The highest number actually has Liechtenstein (727,2), but is not taken here for comparison because of its size.

 

low value of 1.7 in 2014 and the EU cohesion countries are usually varying at higher levels (e.g. Bulgaria 2.1 in 2014; Czech Republic: 13.8; Hungary 23.2; Poland: 3.7; Romania 2.6; Slovakia 8.1).¹¹⁸

 

Finally, the international scientific co-publications per million populations are taken as a proxy for the international openness and connectedness of the domestic research communities with their fellows from abroad. Also with respect to this indicator, Ukraine shows a rather low, but steadily increasing performance ranging from 40.1 in 2005 to 59.4 in 2014. For comparison, Russia shows 76.3 in 2014, Moldova 56.8, Turkey 75.1, Slovakia 372.4, Romania 163.7, Poland 235.2, Hungary 398.1, Czech Republic 610.0 and Bulgaria 175.4. The top performers in 2014 are Switzerland (2,743.2), Iceland (2,364.3), Denmark (1,889.5), Sweden (1,670.2), and Norway (1,527.2).¹¹⁹

 

6.3. Introduction to the bibliometric co-publication analysis ¹²⁰

 

Co-publications are regarded as one indicator for measuring cooperation and are used as one of many proxies for the assessment of the current state of (bi-regional) collaboration in sciences. The following sections assess the activity and impact of Ukrainian research and international research cooperation based on bibliometric findings and discuss recent developments in academic cooperation between EU28/AC and Ukraine and points to emerging scientific topics.

 

The analysis of the publication data of Ukraine is based on the two main academic citation databases Web of Science (WoS) and Scopus for the timeframe 2003 to 2013.

 

It provides an overview on:

 

· the Ukrainian scientific publication and co-publication output,

 

· the most involved partner countries in Ukraine's co-publications,

 

· the main scientific research fields of Ukrainian publications,

 

· overall co-publications and co-publications with EU28/AC¹²¹ countries with special focus on the joint Ukrainian-EU28/AC priority areas Aerospace & Aeronautics, Biotechnology, Nanoscience & Nanotechnology and Information & Communication Technologies,

 

· the Specialisation Index of Ukrainian scientific publications and some highlights regarding scientific impact.

 

Data and methodology

 

The data for the following detailed analysis was retrieved in summer 2014 from Elsevier's Scopus database (Scopus) and Thomson Reuter's Web of Science (WoS). The data cover a 11-year-period from 2003 to 2013 for the overall Ukrainian scientific publication output. On the basis of the retrieved raw data, raw data tables containing records and affiliations from Scopus and WoS were created separately. A combined data set was then created using a series of processing steps in an SQL database and with a specifically developed web interface for a multi-stage data cleaning process (e.g. duplicate detection, raw data correction) including both, automatic and manual steps.

 

The data unification and cleaning steps lead to a significant increase in data quality and a remarkable gain in data coverage. After the unification process 92,763 Ukrainian publications could be identified. 62,376 of these publications were listed in WoS only and 80,335 publications in Scopus only.

 

The method and data used has also certain restrictions:

 

 

118 These data are taken from CWTS.

 

119 These data are taken from CWTS.

 

120 This chapter is mostly based on the Deliverable 2.20, EU-Ukrainian co-publication analysis including emerging trends, funded under grant agreement no 311839 (BILAT-UKR*AINA) by the European Union’s Seventh Framework Programme for research, technological development and demonstration.

 

121 The 28 EU Member States and the countries associated to the EU's Framework Programme 7 (FP7). These include Turkey, Montenegro and Macedonia; Switzerland, Israel, Norway, Iceland, Liechtenstein as well as Western Balkan countries.

 

· Impact measures (cited counts) should be treated with cuation, because data can only give punctual snapshots (August/September 2014 in this study) and the cited counts are naturally constantly changing. Also research fields with a small number of (co-)publications should be treated with caution because low number of records can skew the results.

 

· Control of duplicates: In case a specific piece of research is published via multiple channels in similar ways, there is no way of control for this kind of duplicates at the meta-level.

 

· Limitations due to the general validity of bibliometric data and limitations inherent to the data source (with regards to the amount and coverage of journals and the quality of the data source e.g. misspellings, ambiguity in subject classification etc.) exist and have to be accepted. Despite considerable efforts in data processing and cleaning, there is always a certain margin of error in the data to be considered (a rough analysis of possible errors points to an error probability of 1-5%).

 

· Limitations in benchmarks: The data set is unique and therefore hardly comparable with total sums published in other studies as they usually only use one data source. If benchmarks have to be made, figures by Scopus/SCImago are being used, but direct comparison has to be interpreted with caution.

 

· Comparability of research fields: The average number of authors per (co-)publication is typically significantly higher in some fields (e.g. physics) than in others.

 

 

Cultural and organisational aspects of scientific communities have to be considered as well. For Ukraine, the role of scientific publications in the academic community have been determined by a considerable decline in the number of researchers in the country¹²² and the low incentives for publishing in international journals until 2012 as the system of academic promotion was based on the number of publications in national journals. Additionally, the poor knowledge of foreign languages, especially English, hinders publications in international journals.