The summary statistics for the selected collection of PubmedIDs are shown in Table II. The predicted value is compared to the article wise average number of authors and the average number of times the articles were mentioned. The expected values are the averages of all articles in PubMed that have been matched by journal and year of publication. The “PT” predicted values also influence the kind of publishing. Except as specifically stated, self-citations (an author referring to his or her work) are not included in the analysis.

Multiple publishing categories can be attributed to the same article in Medine/PubMed. Articles are matched across all publishing types for computing “PT” values. For instance, an article with the kinds “Abstract; Multicenter Study; Clinical Trial” will only be compared to other articles with those three kinds. As a result, while evaluating the “PT” numbers, take in mind that there are frequently relatively few publications that match on the journal, year, AND all publishing kinds, which might skew the findings.

For this article, papers are selected from the PubMed database that included the keyword “Social Media” and checked research studies reported in the database between 2000, January and 2021, June.

Pubmed database provides coverage on Medline, dental journal, nursing journal. The PubMed database yielded a total of 713 documents while the term “Social Media” was searched. Even though this research has been conducted in a variety of disciplines, we plan to review the published literature in the Indian context. Only papers with the English language were considered.

713 articles were chosen using a systematic method for social media for health information in India and bibliographical methods, which included a co-citation analysis of reference and text analytics of the combination of bibliometric texts.

To sum up, the keywords reflect that the social media researches on basic health conditions and health communication in India showed the need and relevance for developing countries and rural areas.

Table V shows the number of journals (NumPubs) and the share of the total published publications (percentage of pubs) appear for each journal. Here it is clear that ‘Journal of Family Medicine and Primary Care’ have more publication; but, ratio of average number of citations is highest in ‘Asian Journal of Psychiatry’. The quotes have the same significance as the summary table in general.

Table VI shows year wise number of articles (NumPubs), number of times any article in that year (NumCitesAll) has been cited, inclusive of self-quotation (CumPubs), the number of times no articles have been mentioned in the same year (NumCites); the number of cumulative articles mentioned, including self-quotations (CumCitesAll); the cumulative number of articles cited in that year (CumCitesAL); (CumCites).

The bibliographic analysis of the articles will help to identify research gaps and potential prospects (Xu et al. , 2020). The study of co-quotation analytics and the bibliographic mix of chosen papers are discussed in the following pages.

Identification of co-occurrences, co-authorships based on author keywords are among the research features of the selected 713 articles. This report demonstrates how social media is used in India for health information. It also highlights the most active, most studied fields for potential advancements in the field assessment. The results of the experiments are discussed in the following pages.

The network visualization analysis for the authors’ keywords is seen in Figure 3. To map the results, VOSviewer was used by Van Eck and Waltman (2009, 2013). This software is not only useful for generating, analyzing, and exploring network data maps, but it is also useful for bibliometric data analysis (Van Eck & Waltman, 2013). Perianes-Rodriguez et al. (2016) suggested the fractional counting process and the overall sensitivity of the relation to normalizing the effects. For co-occurrence data, the Van Eck and Waltman data index (Pai & Alathur, 2019) is used.

A total of 727 keywords out of 3105 were obtained for this research study with 713 papers by setting a metric for counting, form, analytical unit, and threshold value as fraction metric, coexistence analysis, and minimum occurrence of keywords as 2.

The size of the circle in Figure 3 reflects the frequency of the keyword. The larger the diameter, the more often the keyword appears in social media for health information journals. The distance between them shows the topic’s relative intensity and similarity (Guo, et al. , 2019). In this report, keywords like humans, social media, India, Covid 19 have a higher weight. The use of the same color in different publications indicated a related subject.

The network visualization map displayed in Figure 3 has 8 clusters that describe the subfields of social media. The Yellow Cluster comprises varieties of family problems and their information such as family planning services, survey, health, postnatal care, risk factors, breast cancer awareness, geography, fertility, health education, family characteristics, etc. The red cluster included social media, online learning, pharmacovigilance, machine learning, natural language processing, algorithms, neural networks, etc. The keywords in the green cluster are human, measles, management, cluster analysis, infant nutrition, qualitative research, adolescence, diet, etc. Next, the keywords in the blue color cluster are correlated with India, social phobia, students, universities, etc.

The purple cluster contains pathology, what’s app, Facebook, etc. keywords which are there. A cluster of other blue colors are combined of Covid 19, content analysis, gender, fake news, sentiment analysis, deep learning, anxiety disorder, information, periodical as a topic, Twitter, etc. The orange cluster, includes key terms basic reproduction number, culture media, circadian rhythm, parents, social media use, adolescents, behavior, students, etc.

By reviewing quotations, this paper examines “the frequency with which two documents are cited together by other documents”(Small, 1973). It is also very important (Figure 4). A total of 91 papers out of 4488 were obtained for this research with 713 papers and the minimum number of papers per author is 2.

This diagram illustrates the interactions between the primary authors and the remaining researchers in the field of social media connected to health information in India. The first cluster, led by Dash, Chinmaya; Gupta, Ravi and Raheja, Amol have the most members. There are a total of 8 writers. The next cluster (green) has 7 writers, the most prominent of one is Kar, Sujita Kumar. Both yellow and blue cluster has the same number of writers; that is 6. For the yellow cluster Banerjee, Debanjan is the most prominent one and in the blue cluster Garg, Kanwaljeet, and Chairasia, Bipin are the most prominent ones. The purple cluster contains 5 writers, where each author has the same importance. In the other blue cluster 4 writers are there and among them Grover, Sandeep is the most prominent one.

A total of 3675 terms out of 20321 were obtained for this study with 713 papers and minimum occurrence of terms as 2. 60% of most relevant terms are selected out of 3675 which is 2205

The size of the circle in Figure 5, reflects the frequency of the keyword. The larger the diameter, the more often the keyword appears in social media for health information journals. The distance between them shows the topic’s relative intensity and similarity (Guo et al. , 2019). In this report, keywords like workers, group discussion, detection, and surgeon have a higher weight. The use of the same color in different publications indicated a related subject.

The network visualization map displayed in Figure 5 has 9 clusters that describe the subfields of social media. The Yellow Cluster comprises college, self-breast examination, medical science, socioeconomic stratum, etc. The red cluster included detection, fake news, reconstruction, processing, computing, modeling, hybrid approach, architecture, cell, gene, generation, etc. The keywords in the green cluster are biopsy, pathologist, mask, specialty, abnormality, coordination, comprehensive strategy, daily activity, etc. Next, the keywords in the blue color cluster are correlated with ultrasonography, internalization, artery intima, cardiovascular disease, origin, glucose, BMI, depressive symptom, etc.

The purple cluster contains young child, infant, IPC, differential impact, program activity, etc. keywords that are there. A cluster of other blue colors are combined of the social media channel, meta-analysis, peer review, citation, altmetric score, e-survey, cloud, bridge, artificial intelligence, etc. The orange cluster includes key terms caste, residence, national health survey, wealth index. Young mother, recent birth, effective implementation, social group, socio-economic group, interval, marital status, logistic regression model, etc. The pink cluster includes keywords like group discussion, media campaign, content analysis, suicide, broader health analysis, important challenge, Chennai, college suicide, media professional, qualitative interview, media guideline, etc. Other green cluster contains Uttar Pradesh, northern India, pediatrics, informal learning, businessman, mobilizer, Hindi, current situation, force, sale, coding, neglect, process, interrupt transmission, etc. keywords. The grey cluster includes ethical aspects, sex selection, couple, live birth, baby boy, release, colony, activity pattern, circadian clock, etc. as the keyword. Other pink colors have surgeon, clinical practice, mainstream, social media channel, variance, main source, the social media network, privilege, agreement, demographical data, chat group, Linkedin, post-graduate training, common degenerative condition, invasive spinal surgery, etc. are there as keywords.

Martinez et al. (2019) in their study have shown that Journals are having largest number of publication a finding which is similar to our study. Like this study it also has ‘humans’ as keyword highest number of times and it is clear in Co-occurrence study. It, Pai & Alathur (2021), Zyoud et al. (2018) and Madjido et al. (2019) also has similar publication trends.

In Co-occurance analysis ‘Telemedicine’ is an important keyword (Pai & Alathur, (2021); Madjido et al. (2019)); as is echoed in our article. However Co-citation analysis in our study is marketly different; perhaps they have worked on M-health.

Zyoud et al. (2018) mentions 10 most active journals which are not a part of our findings. This could be because we have focused on social media where as he has focused on Internet.

The usefulness of mHealth is the other important finding in the clustered papers. MHealth has been used to combine health awareness with phone (Pai & Alathur, 2020).

Social media health information has arrived and will be more prevalent in future as reach of social media extends to developing and under developing countries. The regulators therefore has the challenges to regulate it in such a way that its positive influence remains higher than its negative connotations. Health care Regulators should frame policies so that public health may benefit immensely form the use of social media base health information. Prevention of disease is important and social media can play an important role. As this is an emerging issue academicians and researchers are required to pay greater attention and develop new theories and models for better societal health outcomes. As it appears social media is a greater importance for marginalized people; health care providers and regulators must take precautions to avoid possible negative outcomes.

There are some flaws in the study. First, there were only papers from publications of the same ‘PubMed’ index, with no listing of articles from other sources. Second, the science Only English literature is permitted; no other languages are possible. The total consolidation of the study findings could be harmed (Müller et al. , 2018). Third, “social media” is the keyword for this review. The future researchers in this respect may include databases other than PubMed. Also newer keywords such as social network, social networking sites may be used for deeper analysis. Finally comparisons of two or distinct regions may provide information on differences across geography.