The knowledge domain and emerging trends in apple detection based on NIRS: A scientometric analysis with CiteSpace (1989–2021)

Abstract In this paper, 317 literature in the Web of Science (WoS) related to research on apple by near‐infrared spectroscopy (NIRS) were drawn on the knowledge map of the number of literature, the co‐occurrence network of authors and institutions, the co‐occurrence and clustering of keywords based on CiteSpace. And a related analysis was carried out. Combined with the results of visual analysis and related literature, the research hotspots were sorted out and discussed. This paper provides a certain reference for relevant researchers to study in this field and provides a new method for macroscopically grasping the current status of apple quality detection research, which helps new researchers to quickly integrate into this field and obtain more valuable scientific information.

With the increasing application of NIRS technology in the field of apple detection, and the maturity of technology, it is necessary to synthesize the current state of knowledge and clarify the evolution of this field and its future development direction. The traditional review is mainly based on the induction and summary of relevant literature, sorting out the research results and progress, the research direction is relatively single, only the macroscopic qualitative description and revealing of certain laws and conclusions, although these existing reviews are very valuable for scholars to understand the field. However, they mainly rely on qualitative methods to review the content and themes of the existing literature, and it is difficult to comprehensively and objectively reflect the whole picture of the research field, and it is difficult to systematically show the development process of the research field.
As a research method in the fields of scientometrics and informetrics, a scientific knowledge graph can not only reveal the source of knowledge and the law of development, but also reveal the structural relationship and evolution law of knowledge in related fields in the form of graphical expression . Based on CiteSpace software, this study describes the distribution characteristics of publications, the international collaboration of countries/ regions, the co-occurrence of subject categories, and the evolution of research hotspots using bibliometric and scientometric methods.
These results may help new researchers quickly integrate into the field of apple detection based on NIRS, as they can easily grasp the frontiers of apple detection based on NIRS research and obtain more valuable scientific information. This study also provides references for the follow-up research of relevant researchers.

| Data sources
This article uses the Web of Science (WoS) database as the data source, and the search deadline is December 31, 2021. The WoS database search selects "Web of Science Core Collection," and inputs in the search formula: TS = ('near-infrared spectroscopy' OR 'NIR*') AND TI = (apple), the retrieved results were further screened by language (select English), document type (select review and paper), etc.
With the help of CiteSpace data deduplication function, 317 documents were finally obtained.

| Research methods
CiteSpace software is a visual analysis software based on bibliometrics. It needs to run in the Java environment. It can analyze relevant information in a large number of documents (such as the publication and cooperation of authors and institutions, keyword co-occurrence and clustering, national cooperation, etc.) which is displayed in a visual form, and the relevant information of a certain research field is selectively presented on the map according to our needs, so that researchers can find effective information from it, and intuitively analyze the research development context and hotspots and trends in this research field . This study is based on this software to conduct a review of apple detection research based on NIRS.

| Literature published analysis
The number of published literature is an important indicator for evaluating the development process of this field. Draw a line graph for the number of published literature counted by CiteSpace ( Figure 1).  (Robert et al., 1989).
From 2011 to 2021, the number of published literature showed a trend of significant increase. The number of published literature at this stage accounted for 69.40% of the total. This was mainly due to the significant increase in apple production and the further improvement of people's economic level. More and more attention is paid to the quality of fruits represented by apples. In 2020, the number of published literature reached a maximum of 44. According to statistics, the top three countries in apple production, namely China, the United States, and Turkey, have apple production of 40,501,041, 4,650,684, and 4,300,486 (unit: ton). The above trends in the number of published literature reflect that many scholars are paying more and more attention to research in this field, which is consistent with the development law of the apple industry worldwide, reflecting the increasing demand for apple detection.

| Author cooperation network analysis
The author collaboration network graph can reflect the core authors in the research field and their collaboration and mutual citation relationships (Chu et al., 2022). Based on the author analysis function of CiteSpace, the cooperative network and relevant author information in the fruit detection field are obtained ( Figure 2, Table 1). Each node represents an author, and the size of the node.
The connection between nodes, and the width, respectively, represent the amount of published literature, the cooperation relationship, and the strength between the authors of the published literature. Figure 2 shows a total of 519 and 945 cooperation lines, with a density of 0.007. Some researchers in this field have formed stable teams and cooperated relatively closely. For example,    Table 2).

TA B L E 1 Statistics of top 20 authors on apple detection by nearinfrared spectroscopy (NIRS)
No. Author

| Keyword co-occurrence analysis
Keywords carry the most important and core information of the literature, and are the key to grasping the important information in the literature. Therefore, we can understand the research hotspots in a certain field by analyzing the keywords with high frequency (Zheng et al., 2020). Use CiteSpace to analyze the keywords of apple detection research literature based on NIRS, set the Node Type to Keyword, the threshold to T = 30, and the rest to default.
Eliminate invalid keywords and combine multiple similar keywords, and finally, the research hotspot knowledge map shown in Figure 5 and the top 30 keyword information shown in Table 4 are obtained.
The research hotspots of apple detection based on NIRS technology are preliminarily analyzed based on high-frequency keywords.
These keywords come from the title, abstract, author keywords, and keywords provided by WoS. The size of the nodes indicates the frequency of the keywords, and the connection between the nodes reflects the co-occurrence strength and relationship of the keywords. The larger the node, the higher the frequency of keyword occurrence, and the thickness of the connection line indicates the strength of co-occurrence between keywords. Figure 5 and

| Subject co-occurrence analysis
Subject co-occurrence analysis was performed on all data. Subject co-occurrence maps and statistical tables were drawn ( Figure 6, Table 6). Among them, Food Science & Technology has the largest node, the highest frequency, and an earlier appearance, revealing that NIRS-based apple detection is mainly based on this discipline.
It has a high influence, which may be because the apple detection technology based on NIRS needs to correlate with a certain parameter obtained by the chemical detection method.

| Reference analysis
In bibliometric analysis, co-citation analysis can be used to analyze the composition of the knowledge base of a research field. Table 7 lists   (Anderson & Walsh, 2022;Pourdarbani et al., 2022). That is to say, the spectral reflectance or absorptivity of apples in a certain wavelength

F I G U R E 4 National cooperation network
F I G U R E 5 Keyword co-occurrence network is larger than that of other parts. According to this characteristic, combined with the optical detection device, the nondestructive detection of apple quality can be realized (Qin et al., 2021). etc., have appeared in recent years (Anderson & Walsh, 2022;Bobelyn et al., 2010;Fan et al., 2019;Zhang, Huang, et al., 2022).

| Research on apple's internal quality detection based on NIRS
The internal quality detection of apples mainly includes the detection of chemical components. The chemical components include the detection of soluble solids, acidity, vitamins, starch, etc., of which there are relatively many studies on soluble solids content (SSC) and acidity. SSC and acidity are important factors that affect the taste of apples. Since different ripenesses of apples correspond to different acidity values, acidity detection can also be used to judge the ripeness of apples (Pourdarbani et al., 2022). Table 8 is a case study of apple's internal quality inspection based on NIRS. The practice has proved that for the detection of apple soluble solids and acidity, the prediction model established by PLS has a good effect, and high accuracy, and is widely used.

| Research on apple external quality detection based on NIRS
The external quality detection of apples mainly includes the detection of external damage, rot, and pesticide residues. Damage and rot are due to the peeling or cracking of the skin due to bumping and squeezing during manual picking, handling or transportation, and further development will eventually lead to discoloration and rot (Mogollon et al., 2020;Nturambirwe et al., 2020;Tang et al., 2020). Pesticide residues are mainly caused by the excessive application of pesticides and the use of unreasonable cleaning methods  used PCA based on NIRS to detect apple bruise damage, and the overall detection accuracy was 99.5%. Nturambirwe et al. (2020) successfully not only detected bruises of three apple species based on near-infrared (NIR) imaging technology and partial least squares discriminant analysis (PLS-DA), but also pointed out that the fruit variety has a certain impact on the bruise detection ability.

| Research on apple disease detection based on NIRS
During storage and transportation of apples, various internal diseases such as brown rot, bitter pit, ring rot, moldy core, and watercore may occur (Chang et al., 2020;DeBrouwer et al., 2020;Grabowski, 2021;Sun et al., 2022). Diseases affect the quality of apples, which are often difficult to detect in the early stage of the disease, and some diseases are infectious. Moldy core, which is  Table 9.

F I G U R E 6 Subject co-occurrence network
At present, most of the apple detection work based on NIRS is in the laboratory research stage, and NIRS detection equipment is expensive, which is not conducive to popularization in the general population. Our next step should be to develop a portable detection device based on a mature theoretical basis, which is convenient for fruit farmers and other related staff to use. Fuji Watercore SGS-SNV + LS-SVM Accuracy = 98. 48% Zhang, Wang, et al. (2022) has done more research in this field; (3) China has the largest number of published literature, but the betweenness centrality value of the United States and Belgium is the largest, indicating that the United States and Belgium have high-quality scientific research results and cooperate closely with other countries in this field; (4) keyword cooccurrence, cluster analysis, subject co-occurrence, and reference analysis show that the research hotspots in this field are NIRS-based apple internal quality detection, NIRS-based apple external quality detection, and NIRS-based apple disease detection. According to the visualization analysis results, the hot research on these three aspects is discussed. This literature provides a certain reference for the relevant personnel of apple detection research based on NIRS, which is beneficial to grasp the research status.

ACK N OWLED G M ENTS
We would like to thank and acknowledge the support of Dr. Mingqing Wu and Dr. Zhijiang Li at the Tarim University in discussions and suggestions regarding the structure of this paper.

FU N D I N G I N FO R M ATI O N
This study was supported by the funding from the National Natural Science Foundation of China (11964030); the Open Project of Key

Laboratory of Modern Agricultural Engineering in Colleges and
Universities of the Department of Education of the Autonomous Region (TDNG2021201).

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data of this study are from WoS and can be reflected in the text.