16 Years of Rita - A bibliometric analysis

1.Introduction

Rita is an brazilian scientific magazine that surges in 1989 with the mission to have a regular and strong computer science publication in the Informatics Institute - UFRGS. Since them, lots of changes have in article publication and RITA continues to carry his legacy across time. With a digital format of publication since 2008, having 16 years of article knowledge stored in database, creates a perfect opportunity to make an exploratory analysis to observe many different aspects of computer science publication inside the magazine.

In this study we try to answer the following questions:

• How the number of publications evolve across 15 years of existence and Which years had the highest and lowest number of publications?
• How do different bibliometric indexers vary in assessing the quality and impact of scientific publications in the journal over the last 15 years?
• Which subjects have more views across 15 years?
• Which articles have most views?
• Which authors and institutions have the highest number of published articles?
• What is the variation of the topics published in our magazine?
• What is the geographical distribution of the authors of the articles published in the journal?
• What is the collaboration network between authors and institutions in the journal’s articles?
• Which authors and institutions are the ones with most articles rejected?
• How different was the keywords of the rejected articles compared to the accepted ones?

2. Background and related work

This chapter introduces a fundamental explanation of bibliometrics and its metrics, methods, and applications. After that, we discuss the related work, their contributions, and their limitations.

2.1 - Bibliometric overview

Bibliometric analysis is a popular and rigorous method for exploring and analyzing large volumes of scientific data. It enables us to unpack the evolutionary nuances of a specific field, while shedding light on the emerging areas in that field (Donthu et al., 2021)

For the usage of bibliometric toolbox, we need to met two conditions:

• The scope of the review is too broad (Donthu et al., 2021)
• Dataset is too large for a manual review (Donthu et al., 2021)

RITA has publication records from 2008 to 2023, and we want to use that dataset to evaluate scientific production, research tendencies, and collaborations related to authors who contributed to the journal.

2.2 - Main techniques

Bibliometric analysis consists of two main techniques: Science Mapping and Performance Analysis

Performance analysis examines the contributions of research constituents to a given field (Cobo, López-Herrera, Herrera-Viedma, & Herrera, 2011; Ramos-Rodrígue & Ruíz-Navarro, 2004)

Science mapping examines the relationships between research constituents (Baker, Kumar, & Pandey, 2021; Cobo et al., 2011; Ramos-Rodrígue & Ruíz-Navarro, 2004).

The image below gave us an overview of the metrics presented in this study:

Table with metric description (Donthu et al., 2021)

The main goal of bibliometric analysis is to use quantitative methods so we can observe patterns presented in the database to identify tendencies, frequency, interations, anything related to authors and articles that’ve submit to the journal across the years.

Other methods can be used to evaluate article production, such as Systematic Literature Review or Meta-Analysis. The first is used to summarize and synthesize the findings of existing literature on a research topic or field, while the other summarize the empirical evidence of relationship between variables while uncovering relationships not studied in existing studies (Donthu et al., 2021).

Since the objective is to explore a large dataset of publications of many different topics and aspects of computer science, while observe the evolution of information related to articles and authors, this study opted for the usage of bibliometrical analysis

3. - Pipeline architecture overview

The following image below are the graphical representation to describe the pipeline defined in this proccess:

A briefly overview of each step:

1. The extraction of dataset contain all the information presented in RITA are made and first we made a cleanup of all fields not used for any type of analysis and related to missing data presented in the columns
2. We padronize the text language generating translation of abstracts to english for purposes of word frequency and analysis After that we partition the dataset into two different datasets to avoid merging problems between different types of articles: 2.1. One dataset of published articles - to allow summarizing views and bibliometric indexes calculations 2.2. Another dataset of rejected articles - to explore more aspects on the side of articles that didn’t have seen the light of day
3. Run macros to padronize name of institutions and with basis on this info, locate city and country respective
4. Using the package of nominatimlite to gather latitude and longitude based on the city where the institutions are located
5. Merging views information with the published articles dataset
6. Editing with Google Sheets to merge the remaining data related to title, abstract and submission date
7. A script was created to scrapping IEEE Tessaro and transform them into a dictionary to be used to padronize tokens into keywords and then subsequently used on the word analysis

4. - The Results

4.1 How the number of publications evolve across 15 years of existence and Which years had the highest and lowest number of publications?

Looking only for accepted and rejected results

Total of articles received classified by language

Since we have a lower number of articles in your database related to 2023, we want to use data until 2022 to ensure that we have the significant large scope of dataset for each year

4.2 How do different bibliometric indexers vary in assessing the quality and impact of scientific publications in the journal over the last 15 years?

4.2.1 Indexers related to publications

These indicators included in (Gupta and Hasan 2018) measure the relationship between the number of authors and the number of publications.

a. Annual Growth Rate of Publications - ARG

This indicator measures the percentage change in the number of articles between years (Gupta and Hasan 2018). To calculate AGR, the formula used below:

\[ ARG = (\frac{\#\text{Manuscripts_year}[a] - \#\text{Manuscripts_year}[a - 1]}{\#\text{Manuscripts_year}[a]} \times 100) \]

b. Cumulative Annual Growth Rate - CAGR

It corresponds to the cumulative growth rate of the number of publications over a specific period of time (Kulkanjanapiban and Silwattananusarn 2021). It is calculated as:

\[ \text{CAGR} = \left( \frac{\#\text{Manuscripts_year}_{\text{CUM}}[a]^{\left(\frac{1}{\text{year}(a) - 1}\right)}}{\#\text{Manuscripts_year}[a]} \right) \times 100 \]

c. Relative Growth Rate - RGR

It corresponds to the increase in the number of articles published per unit of time (Kumar and Kaliyaperumal 2015) (Rathika and Thanuskodi 2021). The average RGR for a time interval can be calculated using:

\[ \text{RGR} = \frac{\ln(\#\text{Manuscripts_year}[a]) - \ln(\#\text{Manuscripts_year}[a-1])}{\text{year}(a) - \text{year}(a-1)} \]

d. Doubling Time - DT

There is a direct relationship between RGR and the time it would take for the number of publications to double. To calculate the doubling time (DT), a standard natural logarithm of 0.693 (Kumar and Kaliyaperumal 2015) is used here (Rathika and Thanuskodi 2021), applying the following equation:

\[ \text{DT} = \frac{0.693}{\text{RGR}} \]

Years	Freq	Freq_AC	AGR	CARG	RGR	DT
2008	94	94	0.00	0.00	0.00	0.00
2009	101	195	7.45	93.07	0.73	0.83
2010	64	259	-36.63	101.17	0.28	2.15
2011	30	289	-53.12	112.78	0.11	5.48
2012	60	349	100.00	55.30	0.19	3.17
2013	33	382	-45.00	63.20	0.09	6.70
2014	40	422	21.21	48.10	0.10	6.03
2015	60	482	50.00	34.67	0.13	4.64
2016	32	514	-46.67	41.49	0.06	10.05
2017	50	564	56.25	30.90	0.09	6.70
2018	90	654	80.00	21.94	0.15	4.02
2019	75	729	-16.67	22.97	0.11	5.48
2020	45	774	-40.00	26.75	0.06	10.05
2021	39	813	-13.33	26.32	0.05	12.06
2022	41	854	5.13	24.22	0.05	12.06
2023	18	872	-56.10	29.52	0.02	30.15

4.2.2 Indexes related to the authors

a. Author productivity - AAPP and PPAA

These indicators included in (Gupta and Hasan 2018) measure the relationship between the number of authors and the number of publications.

\[ \text{AAPP} = \frac{\#\text{Authors_year[a]}}{\#\text{Manuscripts_year}[a]} \]

\[ \text{PPAA} = \frac{\#\text{Manuscripts_year}[a]}{\#\text{Authors_year[a]}} \]

b. Degree of Collaboration - DC

It is defined as the relationship between articles with more than one author (Nm) and those written by only one person (Ns) (Savanur and Srikanth 2009) (Gupta and Hasan 2018) (Barik and Jena 2019) (Das, Kaur, and Verma 2021).

\[ \text{DC} = \frac{N_{m}}{N_{m} + N_{s}} \]

c. Collaborative Index - CI

It is a weighted average of authors per joint article. That is, those who have more than one author (Savanur and Srikanth 2009) (Gupta and Hasan 2018) (Barik and Jena 2019) (Das, Kaur, and Verma 2021)

\[ \text{CI} = \frac{\sum_{j=1}^{A} j * f_{j}}{N} \]

d. Collaborative Coefficient - CC

It is a weighted average of authors per joint article. That is, those who have more than one author (Savanur and Srikanth 2009) (Gupta and Hasan 2018) (Barik and Jena 2019) (Das, Kaur, and Verma 2021).

\[ \text{CC} = 1 - \frac{\sum_{j=1}^{A} \frac{1}{j} * f_{j}}{N} \]

Where:

• j is the number of authors per article, example: 1, 2, 3, …A
• fj is the number of articles with j authors.
• N is the total number of articles published in that year.

e. Modified Collaboration Coefficient - MCC

This indicator is applied and described in (Das, Kaur, and Verma 2021).

\[ \text{MCC} = (\frac{N}{N-1})(\frac{\sum_{j=1}^{A} \frac{1}{j} * f_{j}}{N}) \]

Where:

• j is the number of authors per article, example: 1, 2, 3, …A
• fj is the number of articles with j authors.
• N is the total number of articles published in that year.

f. Average number of pages per article (PNA)

Corresponds to the average number of pages per published article.

\[ \text{PNA} = \frac{\sum \text{#Number_pages}} {\text{#Manuscript_by_year}} \]

Years	Freq	AAPP	PPAA	DC	CI	CC	MCC	PNA
2008	94	2.10	0.48	0.76	2.29	0.45	0.46	19.35
2009	101	3.46	0.29	1.00	3.61	0.68	0.71	20.53
2010	64	2.86	0.35	0.77	2.91	0.52	0.54	14.02
2011	30	3.00	0.33	0.80	3.00	0.54	0.60	19.33
2012	60	2.85	0.35	0.95	3.00	0.62	0.66	23.82
2013	33	2.29	0.44	0.79	2.36	0.48	0.52	21.30
2014	40	3.08	0.32	0.92	3.08	0.61	0.66	20.75
2015	60	2.95	0.34	0.85	3.00	0.57	0.60	25.12
2016	32	3.56	0.28	1.00	3.56	0.67	0.75	22.66
2017	50	2.78	0.36	0.83	2.78	0.54	0.57	14.70
2018	90	3.00	0.33	0.90	3.10	0.59	0.61	13.28
2019	75	3.08	0.32	0.75	3.12	0.52	0.54	11.87
2020	45	2.65	0.38	0.88	2.65	0.54	0.58	13.91
2021	39	2.92	0.34	0.85	3.00	0.56	0.61	14.67
2022	41	3.08	0.32	0.85	3.15	0.59	0.64	10.95
2023	18	4.25	0.24	1.00	4.50	0.77	1.02	12.22

4.2.3 Hipotesis proof for average of the indexes

We realize an statistical hipotesis test to see if our sample is significant enough. All tests realized in this section was using a confidence interval of 95% and p-value of 1. The results of each average are in the images down below

Average number of publications

Average number of pages

4.3 Which subjects have more views across 15 years?

Views per year 2022

Views aggregated by Area in 2022

area	views
Business	4239
Electric variables	3444
Artificial intelligence	2104
Computer science	1430
Computational linguistics	369
Algorithms	288
Materials science and technology	267
Circuits	260
Modeling	260
Medical services	214

Views per year 2018

Views aggregated by Area in 2018

area	views
Economics	64704
Computer science	44143
Industries	24660
Business	23400
Optics	19674
Electric variables	16364
Modeling	6572
Pattern recognition	4182
Metals	3852
Algorithms	3301

Views per year 2013

Views aggregated by Area in 2013

area	views
Computer science	40964
Electric variables	21870
Economics	19008
Industries	13080
System analysis and design	4232
Optimization	1842
Modeling	1738
Materials science and technology	872
Image processing	869
Materials	0

Views per year 2008

Views aggregated by Area in 2008

area	views
Artificial intelligence	430288
Electric variables	110438
Robots	101007
Design methodology	86295
Industries	72905
Computer science	44352
Materials science and technology	11124
Information systems	10980
Modeling	9779
Probability	9225

4.4 Which articles have most views?

The 10 articles with the most views

author	title	institution	Views	year
André C. P. L. F. de Carvalho	Uma Introdução às Support Vector Machines	Universidade de São Paulo - USP	51663	2008
Maurício Marengoni	Tutorial: Introdução à Visão Computacional usando OpenCV	Universidade Presbiteriana Mackenzie	27564	2009
Adriana Bordini	QEDS: Um Simulador Clássico para Distinção de Elementos Quântico	Universidade Federal de Pelotas - UFPel	25779	2016
Lino Sanabria	Estudo do Padrão Avançado de Criptografia AES – Advanced Encryption Standard	Universidade Federal da Grande Dourados - UFGD	13048	2011
Paulo Blauth Menezes	Teorias da Aleatoriedade	Universidade Federal do Rio Grande do Sul - UFRGS	9708	2008
Rafael Prikladnicki	Um Olhar Sociotécnico sobre a Engenharia de Software	Universidade Federal do Rio de Janeiro - UFRJ	9613	2008
Jacques Wainer	Processos de Decisão de Markov: um tutorial	Universidade Estadual de Campinas - Unicamp	9516	2008
Siome Klein Goldenstein	A Gentle Introduction to Predictive Filters	Universidade Estadual de Campinas - Unicamp	9225	2008
Rafael Santos	Java Advanced Imaging API: A Tutorial	Instituto de Estudos Avançados – IEAv	8010	2008
Heimar de Fátima Marin	Desenvolvimento e Avaliação de um Protocolo Eletrônico para Atendimento e Monitoramento do Paciente com Doença Celíaca	Universidade Federal de São Paulo - UNIFESP	5946	2010

4.5 Which authors and institutions have the highest number of published articles?

Authors who published the most articles - Added until 2023

Autores	Institution	Frec
Anderson Faustino da Silva	Universidade Estadual de Maringá - UEM	6
José Palazzo Moreira de Oliveira	Universidade Federal do Rio Grande do Sul - UFRGS	6
Aldo von Wangenheim	Universidade Federal de Santa Catarina - UFSC	4
Eanes Torres Pereira	Universidade Federal de Campina Grande - UFCG	4
Luciano Antonio Digiampietri	Universidade de São Paulo - USP	4
Alfredo Silveira Araújo Neto	Techway Informática Ltda	3
Gustavo Rau de Almeida Callou	Universidade Federal Rural de Pernambuco - UFRPE	3
Ivan Torres Pisa	Universidade Federal de São Paulo - UNIFESP	3
Jaime Simão Sichman	Universidade de São Paulo - USP	3
Judith Kelner	Universidade Federal de Pernambuco - UFPE	3

Institutions that published the most articles - Added until 2023

Instituciones	Frec
Universidade de São Paulo - USP	75
Universidade Federal do Rio Grande do Sul - UFRGS	59
Universidade Federal de Pernambuco - UFPE	46
Universidade Federal de Pelotas - UFPel	34
Universidade do Estado de Santa Catarina - UDESC	27
Universidade Estadual de Campinas - Unicamp	26
Universidade Federal Rural de Pernambuco - UFRPE	25
Universidade Federal de Santa Catarina - UFSC	24
Universidade Tecnológica Federal do Paraná – UTFPR	24
Universidade Federal de Juiz de Fora - UFJF	22

4.6 What is the variation of the topics published in our magazine?

Areas and sub-areas of research observed in 2023

Areas and sub-areas of research observed in 2018

Areas and sub-areas of research observed in 2013

Areas and sub-areas of research observed in 2008

4.7 What is the geographical distribution of the authors of the articles published in the journal?

Map with aggregated localization of authors

Authors localized in 2023

Total authors location in 2018

Total authors location in 2013

Total authors location in 2008

4.8 What is the collaboration network between authors and institutions in the journal’s articles?

Aggregated connections of institutions across 15 years of RITA

Relations between authors and institutions in 2008

Relations between authors and institutions in 2013

Relations between authors and institutions in 2018

Relations between authors and institutions in 2023

4.9 Which authors and institutions are the ones with most articles rejected?

Authors

Autores	Institution	Frec
Flávia Gonçalves Fernandes	Universidade Federal de Goiás - UFG	4
Carla Diacui Medeiros Berkenbrock	UDESC	3
Mohanad	Al-hikma University	3
Naidson Clayr Santos Ferreira	Instituto Federal de Educação, Ciência e Tecnologia	3
Alexandra Katiuska Ramos Diaz	Universidade de São Paulo	2
Almir Pereira Guimarães	Federal University of Alagoas	2
Aloysio Augusto Rabello de Carvalho	Universidade Federal de São Paulo	2
Ana Cervigni	Centro de Tecnologia da Informação Renato Archer-CTI	2
Beatriz Terezinha Borsoi	Universidade Tecnológica Federal do Paraná	2
Bharti Bisht	MRIIRS	2

Institutions

Instituciones	Frec
Fundação Educacional de Fernandópolis	12
Universidade Estadual de Maringá	12
Universidade Tecnológica Federal do Paraná	11
Universidade de São Paulo	9
Universidade Federal do Pará	6
Universidade do Vale do Itajaí	5
Universidade Federal de Goiás - UFG	5
Universidade Federal do Rio Grande do Norte	5
Universidade Federal Rural de Pernambuco	5
UTFPR	5

4.10 How different was the keywords of the rejected articles compared to the accpeted ones?

Areas and Sub-areas of study observed in 2023

Areas and Sub-areas of study observed in 2018

Areas and Sub-areas of study observed in 2013

Areas and Sub-areas of study observed in 2008

References

Villalobos, M. (n.d.). Rita: Exploratory Analysis and Visualization of Text. Retrieved June 23, 2024

Barik, Nilaranjan, and Puspanjali Jena. 2019. “Bibliometric Portrait of Select Open Access Journals in the Field of Library and Information Science: A Scopus Based Analysis.” Library Philosophy and Practice 1: 1–8.

Das, Saumen, Kiran Kaur, and Manoj Kumar Verma. 2021. “Publication and Collaboration Pattern of College and Research Libraries Journal During 2009-2018: A Scientometric Analysis.” Library Philosophy and Practice 5118.

Gupta, Simran, and Nabi Hasan. 2018. “Scientometric Analysis of Metamorphosis: A Journal of Management Research.” DESIDOC Journal of Library & Information Technology 38 (4): 254–58. https://doi.org/10.14429/djlit.38.4.12511.

Kulkanjanapiban, Pachisa, and Tipawan Silwattananusarn. 2021. “Bibliometric Analysis of Publications in the Scopus Database: A Study at Prince of Songkla University (PSU) During 1978-2021.” Library Philosophy and Practice 125: 23001.

Kumar, R. Santha, and K. Kaliyaperumal. 2015. “Scientometric Analysis of Global Publication Output in Mobile Technology.” DESIDOC Journal of Library and Information Technology 35 (4): 287–92. https://doi.org/10.14429/djlit.35.4.7884.

Rathika, N., and S. Thanuskodi. 2021. “Studies on Relative Growth Rate and Doubling Time of Publications Productivity of Nuclear Medicine Research.” Journal of Pharmaceutical Research International, 198–211. https://doi.org/10.9734/jpri/2021/v33i32a31732.

Savanur, Kiran, and R. Srikanth. 2009. “Modified Collaborative Coefficient: A New Measure for Quantifying the Degree of Research Collaboration.” Scientometrics 84 (2): 365–71. https://doi.org/10.1007/s11192-009-0100-4.