Introduction

The issue of temple culture has received a lot of attention internationally [1]. Since the 2003 United Nations Convention for the Safeguarding of Intangible Cultural Heritage [2], cultural restoration has gained widespread acceptance among people [3, 4]. This convention emphasizes the significance of the preservation of the "spirit of place," which contributes to the creation of emotional associations and consciousness bonds for an independent region and nation-state and is a key mechanism for the building of national strength. However, the thorough research up to this point has prompted us to reconsider some of the exogenous elements affecting clan cultural transmission, specifically the artistic presentation of ritual in the architectural decoration of clan shrines, which sets it apart from DeLyser and Greenstein's definition of cultural restoration (a relationship of material agency assumed by the clan shrine) [5].

Ancestral halls are challenged by cultural uniformity and loss as a result of urbanization [6]. The Guangdong government will pledge to restore around 1500 ancestral halls by 2020. However, it appears that the cultural restoration is driven more by the desire to change the site of ancestor worship and devotion into a more contemporary cultural production than it is by the need to preserve and transmit the culture of the clan shrines. The ritual economy of worship has been considered a byproduct of economic growth and an effective instrument for earning revenue [7, 8], while tourism has contributed to the multiplication and formalization of clan functions as well as the commercialization of religious activities and artifacts. There are currently 130,000 religious sites that serve as tourist attractions [9]. A strong desire for cultural authenticity is what drives the process of colorizing religious practices [10], and the essence of experiencing traditions should serve as the foundation for cultural transmission.

Ancestral halls were once thought of as essential tools of dominance [11]. Ancestral hall murals are a significant medium for presenting the materialization of the spirit of the past society, and in modern times, ancestral halls have emerged as cultural locations for exploring the social landscape of the past [12]. Mural painting takes on the responsibility of bridging the gap between the past and the present [13], and it does it by directly and traditionally expressing cultural values and societal ideals to the observer. The study of ancestral hall murals may help to uncover the formation of patriarchal relations behind the clan culture, and a few scholars such as Yan [14], Vucetic [15], Rolston [16], Źrałka [17], and Li [18] have paid attention to this aspect.

Yan [14] demonstrated how the mural art was relevant to the society behind it in his analysis of the murals at the Chen Family Ancestral Hall in Guangzhou. According to Vucetic [15], the development of ornamental mural painting in Colombia throughout the ancient time might shed light on the origins of current political connections. In his analysis of the history surrounding the mural’s production, Rolston [16] pointed out several disparities in Chile in 1946 concerning national income, housing, and education. Źrałka [17] countered that the fictional elements of the frescoes might represent the artist’s or the audience's desire for something beautiful, as Hu [19] demonstrated concerning the Guangzhou fresco "Fisherman, Woodcutter, and Cultivator," whose decorative elements and artwork frequently reflected the cultural identity of the area or nation and also had an educational purpose. Li [18] also highlighted the impact of painting schools and skills on the styles of mural painting.

As a result, current research reflects the cultural characteristics reflected in the ornamental art of shrine murals as well as the underlying principles of social evolution, such as elements of national history, geopolitics, and cultural geography. Right now. The rituals that shaped the Guangdong tradition of ancestral hall mural decorating have not been thoroughly studied. In actuality, ritual analysis is among the most archaic aspects of traditional Chinese culture and one of the foundations of the country's power [1]. There is still more effort to be done to document Guangzhou’s architectural murals, of which 4097 have been registered [18]. Several related works [20, 21] have provided detailed descriptions of the content of these murals, enabling additional investigation and serving as an essential foundation for our study.

This study aims to employ text mining techniques to elucidate the pertinent factors impacting the decorative art of ancestral hall murals in Guangdong and to unearth the manifestations of traditional ritualistic elements therein, thereby revealing the intrinsic correlation and interplay between the two, which imparts insights for the preservation, restoration, and cultural perpetuation of both tangible and intangible heritage. KH Coder, an emergent text mining tool conceived in 2014 [22], has been extensively deployed in domestic Japanese research endeavors, predominantly published in Japanese [23,24,25,26]. To date, only a scattering of English expositions elucidates the utilization of this tool [27, 28]. As a modality within the domain of computer science, the application of text mining in the cultural and artistic sphere has yielded scant results [29, 30]. This research proffers an innovative methodological framework and scientific grounding for the utilization of text mining to analyze unstructured artistic data, bridging computer science with the humanities and arts, thereby forging a novel avenue for the integrative analysis of culture and decorative art.

First, the chosen mural materials are manually sorted and classified; next, all the textual materials introducing mural paintings are chosen; next, a text mining algorithm should be used to elicit the relevant influencing factors of the decorative art of shrine mural paintings; finally, a correlation with patriarchal ritual culture is built using thematic clustering and cross-tabulation analysis. Finally, to understand the cultural formation linkages underlying the ornamental art of frescoes, the content of frescoes with an emphasis on figure topics is evaluated to determine influencing elements, and the results are remapped with socio-historical and cultural occurrences.

Research methodology

Data source and processing

The research material for this study originates from "Guangzhou Ancestral Hall Murals", volumes one and two, published in 2015 by Chinese scholar Liu Zhaojiang professor [20, 21]. The initial intention behind the creation of these two books was to systematically comb through and study the unique mural art found in ancestral halls in the Guangzhou region. As an art form that is both irreplaceable and susceptible to damage, murals must be documented and disseminated in forms such as books. With the objectives of cultural preservation, academic research, and educational popularization, Liu Zhaojiang meticulously organized 521 artistically valuable ancestral hall murals. He cataloged the origins, content, artistic techniques, and characteristics of the murals, as well as their relationships to the ancestral halls they reside in, all in a form that combines text and imagery.

Only descriptive information about the murals will be extracted for analysis to concentrate on the theme. Information from the books, such as pictures, authors, dates of inscription, painting names, ancestral hall locations, and scroll specifications, will not be taken into account in the textual analysis stage. The team has compiled this data for analysis into an Excel document containing 521 groups of cell text before extraction. Second, in preparation for the ensuing coding study, the mural subjects and tale contents were divided into external variables Group A and Group B, and new list columns were created by information filtering. The number of murals in groups A and B after coding is shown in Fig. 1 to demonstrate the impartiality of the coding procedure. Table 1 lists the original categorization of subjects and materials that were coded.

Fig. 1
figure 1

Number of murals in groups A and B after selective coding

Full size image
Table 1 Initial coding table for external variables group A and group B
Full size table

Research techniques and procedures

KH Coder, developed by Kosuke Akaishi’s team at Osaka University in Japan, is a text mining tool grounded on artificial intelligence algorithms [31]. This open-source software, purpose-built for quantitative content analysis, provides a suite of functionalities encompassing word frequency analysis, cross-tabulation, co-occurrence network, and multivariate analysis. Rooted in Natural Language Processing (NLP), KH Coder utilizes the R language algorithm [32] and MySQL [33] for data analysis. To date, it has been widely employed in over 900 projects for text collection [31] and quantitative analysis [34], spanning diverse domains such as linguistics [22], anthropology [35], economics [36], and sociology [37]. The software's strength lies in its objectivity. Its unstructured, robust analytical functionality is particularly adept at uncovering trends and characteristics [27].

Compared to ROST CM6 [38] created by Wuhan University in China, KH Coder boasts superior automatic word segmentation and word combination extension capabilities [31]. KH Coder supports multiple language analyses, including Chinese, English, Japanese, and Korean. Functions such as text segmentation, vocabulary integration, and knowledge graph construction are integrated into a single operating panel, making it user-friendly [32]. In comparison to a series of language processing toolkits developed based on the Python programming language by Wuhan University, which requires format conversion for compatibility (e.g., HanLP, THUCTC, Snownlp), and only supports Chinese language, KH Coder presents a more advanced feature [39]. When performing word searches, KH Coder can trace back to all other variants of a term and locate its occurrence in the text, whereas Wordstat, developed by Provalis Research in Canada, can only retrieve the terms entered in the “search bar” [40].

This paper analyzes based on the below steps:

In the first step, Excel was loaded into KH Coder in the form of Stanford POS Tagger to pre-process the information, and check the merging of words in the word frequency list. The nonsense words, such as Be verbs, pronouns, and conjunctions, were then coded into “Force Ignore,” [41] while the semantics of other words were simultaneously checked and the words with fixed collocations were extracted and coded into “Force Pick up,” which was then run again. Following processing, the basic descriptive information was collected. Based on the procedure stated above, the filtered text data was made more legitimate by verifying the word frequency list four times and updating the “Force Ignore” and “Force Pick up” word lists, and the calculation’s accuracy was enhanced (running speed from 53 to 48 s). When compared to the initial information processing results (Table 3), 3509 nonsensical words and 105 classes of related terms were effectively eliminated. The total number of sentences, paragraphs, and cells was 1273, 1273, and 523, respectively, after the words and word classes to be studied were efficiently decreased by 19,508 words and 739 classes with the use of KH Coder's enhanced word processing feature. Figure 2 depicts the results of the four information processing procedures' optimization.

Fig. 2
figure 2

Comparison of the results of the four data processing processed

Full size image

In the second step, we calculate Term Frequency (TF) and Document Frequency (DF). Term Frequency (TF) represents the frequency of a given word's appearance in a document. A high TF is generally considered to be highly correlated with the document’s theme. On the other hand, DF indicates the quantity of a given word in different analytical units, such as sentences, paragraphs, and cells. If a word only appears in a few documents, it could potentially be a key words for describing the themes of these documents. By employing Eq. 1 [42] and Eq. 2 [43], TF and DF are determined, respectively. As illustrated in Fig. 3, there are approximately 7056 terms with a TF < 10. These infrequently appearing terms, which constitute 95.57% of the data, are not adequately representative of the text's value, hence are not included as main analysis targets. The analysis focus is confined to terms within the range of 10 ≤ TF ≤ 100, designated as “Priority” level for subsequent study. It is also noteworthy that terms with exceedingly high TF may not necessarily be meaningful as they are mostly common words and may not be entirely characteristic or representative [44]. Consequently, terms with 100 < TF < 200 are designated as “Inspect” level, requiring semantic verification in subsequent research. There are 454 terms with TF > 200, typically closely related to the text theme but often appear in pronoun form and do not provide meaningful analysis, thus also excluded from the primary analysis target, making up 0.01% of the total. Therefore, by eliminating terms with exceedingly low or high TF, the total analysis workload is effectively reduced by 95.58%.

Fig. 3
figure 3

Calculation results of TF and DF

Full size image

The TF x DF graph depicts the degree of positive correlation between the two metrics [45]. Statistics show that there are 371 terms with TF ≥ 10. Due to space constraints, this paper only displays the top 100 terms, with a total TF of 3972. Nouns account for 60% of the data, verbs account for 31%, and adjectives make up 9% (Fig. 4).

Fig. 4
figure 4

Content data statistics for different impact element groups

Full size image
$$TF(i)=\frac{{\mathrm{log}}_{2}(\mathrm{Freq}(i,j))+1}{{\mathrm{log}}_{2}(L)}$$
(1)
$${\mathrm{idf}}_{\mathrm{i}}=\mathrm{log}\frac{|D|}{\left|\left\{d:d\ni {t}_{i}\right\}\right|}$$
(2)

Word frequency statistics and topic coding

In the third phase, the data were normalised by weight conversion using (Eq. 3) [46], and the key influencing elements were elicited and their affective inclinations were determined using a word frequency descending table, in order to understand the social influence they brought. The frequency of word an in a data source is denoted as \({N}_{a}\), and the score \({N}_{a}^{\mathrm{^{\prime}}}\) is calculated by converting the highest \({N}_{\mathrm{max}}\) of word frequency in the data source that corresponds to a. Finally, a TF ≥ 5 weighted statistics table is obtained (Table 2). Table 3 was created after the words in Table 2 were divided into groups based on their lexical characteristics and coded based on the semantics of the terms in the groupings.

Table 2 Feature words with frequency top 100
Full size table
Table 3 Semantic classification of words in Table 2
Full size table
$${N}_{a}^{\mathrm{^{\prime}}}={N}_{a}\cdot 100/{N}_{\mathrm{max}}$$
(3)

Cross-tabulation and correlation construction

The fourth stage examines the possible influence link between topic groups by counting the text proportion for each subject group in Table 3 according to coding standards and comparing the numbers. The subject groupings with greater weights typically exhibit better correlations within the same text data. The uncompiled text will be categorized as “Other,” and the word groupings without a trend or feature will not be discussed in this paper's percentage graph.

The statistics (Fig. 4) show that among the classified emotional elements, “character” (24.95%) has a greater proportion and that there is a strong link between it and “action verbs” (4.72%). It is feasible to conclude that the murals largely represent positive content when paired with the emotional values in Table 6, with the “Life Allegory” (5.03%), “Fame and Fortune” (4.72%), and “Health” (2.53%) components reflecting the pleasant things that people want in their life [47]. There are two groups for these great sights. The first kind is the pursuit of immaterial things like “Poem (诗),” “Moral (道德),” “Life (生活),” and other phrases that describe the literati and affluent people's way of life. The desire for material life is the opposite, as expressed by phrases like “Welfare (福利),” “Rich (富有),” “Longevous (长寿),” “Lucky (幸运),” and “Beatific (幸福),” which all allude to the pursuit of fame and fortune [48]. Because Guangdong has long been a commercial city, it may be inferred from comparisons that people in ancient Guangdong were more keen to pursue fame and money [49]. Additionally, while “Plants and Animals” (2.92%) and “Artifacts” (4.15%) occupy low weights, “Non-action Verbs” (9.73%) occupy a medium degree of influence, reflecting the indirect and metaphorical logical relationship of information transmission of the mural content [50]. It can be assumed that the decoration of flowers, plants, and household utensils in the mural may play an allegory [51].

In conclusion, three themes—figure form and rhyme, artifact metaphor, and value transmission—can be widely categorized as ceremonial influences on the art of Guangdong ancestral hall paintings. Due to space constraints, this study concentrates on the primary topic of 318 murals, which primarily depict characters; the other two themes will be covered in a later set of studies.

The ritual component of the decorative art of Guangdong ancestral hall murals

The evolution of sitting and lying reflected in the figure's form and rhyme—reflecting the dynamic change of etiquette from respectful to natural

The change in the sitting and lying of the figures in the mural—from formal to relaxed ritual dynamics, to a certain extent reflects the dilution of patriarchal ritual thinking, reflecting the popular aspects of mural art, driven by the policy direction. The combined sitting posture (one leg straight and the other bent at the knee or in a frontal or lateral shape) [52] is depicted in 251 (79%) of the 318 ancestral hall murals (Table 4), whereas the traditional kneeling posture (serious and respectful monarch and minister etiquette) [53] is less common. This is quite different from the dignified sitting posture that was emphasized by traditional rituals in the past. Policy guidance may have been one of the important factors contributing to this change in dynamics. The earliest appearance of the figures seated on pedestals and in the sitting position can be traced back to Ming dynasty murals of ancestral halls [21], and in 1536 Ming Jiajing decreed that “all folk are allowed to establish temples” [54], which led to a boom in the construction of ancestral halls and the appearance of urban ancestral halls in the public eye. The pursuit of freedom and open-mindedness in the city, as well as the cultural influence of emancipation promoted by “Western learning” [55], made it difficult for the traditional stereotypical fresco figures to satisfy the aesthetic needs of the public, which led to the diversification of fresco figures.

Table 4 Changes in the sitting and lying movements of figures in the murals of Guangdong ancestral halls
Full size table

Figure 5 displays, for 315 of the figure murals (excluding the three murals with lacking dimensional data), the relationship between the scale of the painting and the dynamic postures of the figures in the painting. Four characteristics are reflected in the figure: (1) The width of these murals averages 100 cm, and the vast majority of them have a length-to-width ratio of more than 2:1; (2) The proportions of about 15 frescoes are on the square side, and most of the figures in these frescoes are in a standing position; (3) Murals larger than 2:1 in length almost always have non-standing figures in pose; (4) The bulk of the 1318 painted figures are in the standing posture and frequently appear in conjunction with seated and reclining individuals, with a minor number of crouching and kneeling figures.

Fig. 5
figure 5

Proportional analysis of the painting scale and statistics of the number of seated positions of the wall paintings in Guangdong ancestral halls

Full size image

The effect of speedy and efficient artistic creativity spurred by the necessity to accomplish a large assignment may be the shift of the painter from a restrained to a loose brush technique. To outline a vivid and aesthetically harmonious picture full of figures within the limited space of the painting, the lower half of the figure is treated as seated and reclining, which can effectively reduce the imbalance brought by the vertical proportion of the figure to the composition of the long scroll, and this may be a possible explanation for the presence of almost all of the figures in the nearly 200 3:1 long scroll wall paintings in such postures as crouching and reclining. Due to the scarcity of relevant historical information, more evidence is still needed to prove this point.

The ethics of the rules and regulations embodied in the use of colors in costumes—reflecting the change of thought from feudalism to openness

To some extent, the change in the color performance of the characters’ costumes in the mural reflects the social changes in Guangdong in the nineteenth century, from the ethics of the program to the society's open-mindedness, primarily in the use of color from positive color performance to inter-color performance.

The positive color refers to the five pure colors of blue, red, yellow, white, and black (as well as additional pure colors that do not blend) [56], while the inter-color refers to the color that is blended with several positive colors. There were severe ceremonial limits on the usage of colors in ancient China, and positive colors and inter-colors were instruments to distinguish the noble and humble and were not to be blended in any way [57]. The color was also employed to denote authority and position in official attire [58]. Purple was once thought to be a lowly hue as an inter-color. According to Confucius, Garments characterized by an amalgamation of red and purple hues were exclusively designated for the attire of the common populace [57], reflecting the society’s concept of self-control and class differentiation.

In terms of color expression (Table 5), 181 murals (57% of the 318 murals) are represented via positive colors, whereas 137 (43% are expressed through intercoolers). In terms of color expression representing the connection between figures, 149 murals (46.8%) depict the order of inferiority, while 169 (53.2%) show the condition of nature and equality. The quantitative statistics show that the frescoed figures are essentially equal in their portrayal of both contents, but when we examine the time context, this does not match the Ming Dynasty’s clothing color standards [59]. However, the content of the murals may not be a complete representation of social phenomena, since the content of paintings is generally a side reflection of the creator's inner ideas and goals [60], and may also reflect a portion of societal trends [61].

Table 5 Comparison of the color scheme of the mural scrolls and characters in the Guangdong ancestral hall
Full size table

A large number of positive colors such as blue, white, yellow, and black are used to represent the figures in the 318 murals, but some of the murals also portray the figures brightly and subtly, with red and yellow hidden and glowing through the colorful, harmonious, and soft intercoolers—in contrast to the dark, serious, and restrained figures shown in positive colors. These inter-color representations of frescoed figures reflect the fact that the painters at the time were not heavily influenced by class feudalism in their paintings, and the rounded figures also reflect the central idea of human-centeredness and more affirmation of human value and dignity, reflecting the social change of traditional thought.

Perceptual evaluation of murals contents based on semantic differential method

In terms of artistic representations, the textual analysis derived the following results: the subject matter, allegory and content delivery of the Guangdong Ancestral Hall Architectural Murals mostly showed positivity. To validate the above results, this paper will use the Semantic Differential (SD) Method and the Delphi method for perceptual assessment of Mural Content.

Constructing evaluation scales

The semantic Differential (SD) method [62] (also known as the Sensory Recording Method), is a psychometric method proposed by Osgood in 1957, which aims at the psycho-perceptual measurement of subjects’ speech scales. In this paper, the SD method will be taken to investigate the perceptual evaluation of 318 figure murals. The paper will then be evaluated on five dimensions (Color Scheme, Brightness, Composition, Character Depiction, and Theme), and the evaluation process involves constructing scales and asking participants to complete surveys. Referring to the evaluation form of the Likert Scale, authors assigned a more relevant -Negative Adjective Pair to each mural, and participants were required to make evaluations based on psychological feelings. The evaluation is divided into five indicators (with different scores set for each indicator), namely Close to positive adjectives-5 Score, Comparatively Closer to positive adjectives-4 Score, Neutral-3 Score, Comparatively Closer to negative adjectives-2 Score, Close to negative adjectives-1 Score. Table 6 shows an example of an evaluation table for Murals Contents.

Table 6 Examples of perception evaluation forms for murals contents
Full size table

Evaluation process

Drawing on the Delphi method [63], this study invited nine experts in related fields to appreciate and evaluate a sample of 318 murals. The subjects' professional fields are art history scholars, art anthropology professors, postdoctoral fellows in tourism architecture, visual communication designers, environmental design and color research scholars, Ph.D. in Chinese language and literature, and Ph.D. in fusion design. These experts have expertise in history, art history, anthropology, color psychology, and environmental psychology, making the assessment results more scientific.

The evaluation process was divided into three steps. In the first step, basic information such as the purpose of the evaluation, the background of the study, and the data sources were explained to the subjects the use of the SD questionnaire was shown, and then the experts were guided to complete the questionnaire. In the second step, after all the questionnaires were collected, the evaluation opinions were counted and the variance of the Average Score for each Evaluation Indicator of murals was calculated. A variance greater than 1.5 was used as the threshold for divergent opinions. When the variance is greater than 1.5, it means that the experts' opinions are inconsistent, while when the value is less than 1.5, it means that the experts' opinions are consistent. In the third step, the experts were invited to conduct a second round of evaluation on the evaluation indicators with inconsistent opinions and identify the evaluation indicators with variance greater than 1.5. Then, communicate with the experts who gave the evaluation extremes and understand the reasons for the results. This process avoids generating extreme values as much as possible and makes the statistical results consistent [64]. Table 7 counts the adjective’s word pairs. That have appeared in the SD scale.

Table 7 Adjective word pairs that appeared in the SD scale
Full size table

Evaluation results

First, all the Evaluation Indicators made by 9 experts were counted and the Average Score was calculated, the Average Score was counted for the Evaluation Indicators in each category of murals. Table 8 shows the statistical results. With the help of the SD scale, the range of positive and negative score division is clarified and the Delphi method helps to focus the expert opinion. When the Average Score is greater than 3, it can represent that the delivery content of that type of mural is positive in some dimension [65].

Table 8 Calculation of evaluation indicators
Full size table

In Fig. 6, the Average Score of the five categories of FIGURE MURALS for the five Evaluation Indicators can be seen. Historical Allusion's Character Depiction has an Average Score of less than 3 (within the blue pentagon), representing the negative nature of the content being communicated. Correspondingly, all other Indicators have Average Scores in the range of 3–4.5. It represents that the content delivered has a positive direction. The results of this analysis validate that the content of Guangdong Ancestral Hall Architectural Figures Mural Painting generally conveys positive emotional value.

Fig. 6
figure 6

Radar chart of average score for figure mural evaluation metrics

Full size image

Discussions

As a vehicle for visual narratives, murals function as Educationality, Familiality, and Sociability. Table 9 highlights the consensus of relevant studies: (1) Murals are not only an expression of art but also a symbol of culture, history, and identity; (2) Murals can be used as a tool for community development, identity, and cultural heritage education; (3) Murals contribute to public awareness of environmental sustainability, cultural awareness, and a sense of community belonging; (4) The creation and presence of Murals fosters communication and unity in the community. Overall, these studies collectively reveal the centrality of Murals in community building, cultural heritage preservation, and social exchange, emphasizing their combined value in the dimensions of education, Familiality, and Sociability.

Table 9 Functional properties study of ancestral hall murals
Full size table

In addition to the granting of knowledge, modern art education focuses on the transmission of cultural values [83]. In the digital age, traditional art forms are at risk of being marginalized as they appear weak in comparison with the attractiveness of modern technology and media [84]. How to balance globalization and localization to ensure that cultural diversity and uniqueness are preserved deserves in-depth exploration. As a visual medium, the existence and continuity of mural art transcends time and space, conveying the message or values of a particular era. Therefore, murals are not only art; they also carry the memory of history, culture, and society. When traditional elements in murals are gradually lost or diluted, the culture and values they represent suffer [85]. This phenomenon is caused by society's excessive pursuit of instant gratification and fast culture, which in turn tends to overlook deep and enduring cultural values [86], and may ultimately weaken the sense of social identity and affect the stability of the society [87]. According to Sakip et al. [88] art murals are suitable for use in urban backstreets and alleys because they can attract people to use these spaces, create natural surveillance reduce opportunities for criminal activity, and contribute to social cohesion. Art and education are not mutually exclusive. On the contrary, they can complement each other and together contribute to social and cultural development [89]. To this end, we need to revisit the goals and methods of arts education to ensure that they are aligned with the needs and challenges of contemporary society, as emphasized in the study by Ferrer et al. [90].

At the level of theoretical contributions, Vigerland and Borg [91] discuss how cultural capital shapes and influences consumer perceptions in commodity art forms by introducing the theory of “cultural capital”, while this paper specifically focuses on the correlation between economic costs and the dynamic characteristics of mural figures, revealing the diverse socio-economic manifestations of art commodities. By including the social structure and cultural production of a society as the object of study, Suleyeva et al. [92] analyze how rock mural art is constructed and transmitted across the boundaries of time and space from a sociological dimension. In contrast, Prasiasa [93] argues that the traditional experience of art, the perception of value, is crucial for place-making and cultural enrichment. He introduces the concept of “social interaction” to the theory of the public sphere and discusses how mural art communicates with the public and dialogues with modern society from a hermeneutical perspective. This paper provides a unique perspective on this form of artistic communication by specifying this dialog of the times in terms of statistical information on the number of female murals and the dynamic poses of the figures. In the interpretive-historical research dimension, Tansem [94] argues that religious artistic heritage is not only a reproduction of the past but also a critical reading and reflection on historical and cultural traditions, providing a bridge to assess and understand the past and the present. The finding that the color representation of mural figures in the Ming and Qing dynasties differs significantly from folk color systems challenges traditional perceptions and provides empirical support for the practical application of Tansem’s critical framework. In the field of design, the study of traditional artifacts is often centered on the interaction between form and content of the "object" [95], while in the field of phenomenology, the study of tangible heritage often rises to the level of visual language and abstract social meaning [96]. In the field of semiotics, the cultural formative factors behind man-made artifacts are often in focus [97]. Ingram [98] extends the exploration of the formation, emergence, and transformation of traditional art to include a geopolitical dimension, arguing that the origin and form of artworks are the result of a confluence of factors. This paper returns to the exploration of the roots that influence the form of an artwork to its material heritage itself, expanding on the shaping and influence of clan ideology on the art of ancestral hall murals. At the same time, with the help of the KH Coder, this paper provides a concrete path for the application of text mining to art research.

In the method realization path, Berkes et al. [99] presented the notion of “socio-ecological resilience”, which combines the connection between human and social environment as a concept of cultural values to address the process of cultural restoration. By merging the human landscape of locations into the notion of cultural restoration from a historical viewpoint, Zedler and Stevens [100] established the concept of “feeling of place” to examine the variations in cultural restoration orientation between China and the West. Abbasi et al.’s [101] idea of “memory of place” seeks to expand cultural education to the spirit of place to address the purpose of cultural restoration. To offer the meaning of cultural restoration in the modern period, Lapsley [102] introduced the notion of “personality development,” which includes the truth, goodness, and beauty of human nature and the proper values of the subject into the purpose of cultural restoration. Lucas’s [103] notion of “spiritual empowerment” seeks to instill the importance of storytelling throughout the cultural restoration process. This study explores the feasibility of aesthetic education in environmental art through artistic means, revealing the new-age significance of cultural restoration. The research on the influencing factors of rituals is significant because it has the potential to provide new ideas for the transmission of implicit values in cultural attributes. This study throws light on the realization of passing on the excellent cultural genes in traditional rituals.

Figure 7 shows the academic contribution of this study in the field. In terms of theoretical and subject area contributions (Fig. 7a), Scholars in the field of art have developed different research foci around the topic of “tangible heritage culture and art”: (1) for artistry, artists are often concerned with the aesthetic value of their work, their technique and the evolution of their style; (2) for functionality, practicality, innovation and human–computer interaction are universal topics explored by design scholars; (3) for culture, anthropologists, historians, and cultural scientists tend to be more concerned with the restoration, preservation, and transmission of cultural heritages and how they reflect regional, cultural, and national specificities. In terms of methodology and realization path contributions (Fig. 7b), The current research constitutes three different dimensions, and has undergone a thematic expansion from “form and content” → “symbolic elements” → “cultural phenomena”, extending the depth to the level of social meaning. Material artistic expression was an early scholarly concern, with Bell-Villada [104], Gathercole [105], and Welch [106] focusing on the materiality and formal beauty of the artwork. Due to the diversity of the arts and the complexity of the mechanisms of material shaping, this element is then expanded into the fields of semiotics [107], sociology [108], and phenomenology [108, 109]. Carrabine [110], for example, explores how the turn of narrative in criminology can help explain fresco images, arguing that images of extreme violence do not fully reflect the reality of medieval life. Recently, the term “material culture” has been emphasized, and some scholars, such as Zubieta [111], Douglass and Badham [112], and Arroyo-Lemus [109], have argued that works of art are not only the product of aesthetic needs, but also the result of multiculturalism and artisanal skills. At the same time, Chicoine [113] returns to the origins of visual symbols and their decorative elements and builds bridges between art, design, semiotics, and phenomenology, emphasizing the social nature of the tangible heritage itself, and arguing that the study of traditional art should not only focus on the form and technique, but also delve deeper into the cultural and historical connotations behind it. Based on the text-mining technology of computer algorithms, this paper analyzes the text of the mural paintings of ancestral halls with the help of the KH Coder tool and carries out multidisciplinary research that integrates art, design, phenomenology, and semiotics. Compared with traditional single-disciplinary perspectives and research methods, this paper employs information technology tools to provide a wide range of perspectives, frameworks, and examples from the perspective of cultural and ritual criticism, and to ponder how to integrate art and technology to perpetuate and pass on the excellent genes in traditional art. The integration of art and information technology will bring opportunities to traditional art research, as exemplified by Kouretsis et al. [114], Xie [115] and Liu et al. [116], and others. This paper promotes interdisciplinarity in the arts.

Fig. 7
figure 7

Academic contribution map

Full size image

Sippola et al. [117] state that the essence of historical research is to reveal the cultural capital and power relations in the social structure. Stainforth [118] argues that traditional art research is to tap into the collective memory and identity of society. In the sociological dimension, Hallström [119], on the other hand, states that the past is studied to understand the cohesion and structure of modern societies to predict and guide future social change and development. This study draws out the relevant factors affecting the decorative art of murals, explores how traditional patriarchal rituals influence the values of decorative art, and reveals the intertwined relationship between power, identity, and culture in artistic expression, and how to perpetuate the good genes therein in modern society. From the perspective of cultural hermeneutics, the culture and values embedded in the murals themselves have a potential role in promoting social progress [120]. In the context of globalization, the question of how to balance tradition and modernity and preserve cultural diversity and uniqueness has become an urgent one. In terms of ritual education, the revelation of the value of mural art in ancestral halls has a potential motivational mechanism for social promotion and provides a new perspective on artistic empowerment.

The coverage aspect of the mural paintings taken limits the scope of this work. The majority of the paintings (51.6%), 269, were done in the ancestral hall entryway, while the remainder were decorated in the chancel, rear hall, and other sections. Mural content is frequently connected to the location and spatial purpose of their decorating, and a non-equal fraction of sample coverage may lead to partial bias in the study results. To ensure the scientific validity and objectivity of further in-depth research, the researcher should improve in the following areas: (1) authors should select representative samples in a balanced manner, paying attention to the relationship between the location of the mural and its pictorial content to ensure that all parts of the ancestral halls are adequately considered; (2) Consider introducing multivariate data analysis to more fully assess the impact of the cultural, historical, and social context of murals on their content and form; (3) Conduct cross-cultural or cross-regional comparative studies to reveal common features and differences of murals in different cultural and geographical contexts.

Conclusion

The study process showed that: (1) The 521 frescoes' tale themes and allegorical interpretations mainly offer positive content, and the paintings' characters, objects, flowers, and other ornamental art references express the idealistic goal of both material life and spiritual excellence; (2) The change from formal to relaxed sitting and grooming of the figures in the Guangdong ancestral hall murals reflects, to a certain extent, the dilution of the idea of patriarchal propriety; (3) The influence of economic costs may be a possible explanation for the proliferation of figures in crouching, skirting, and side-lying postures in Guangdong ancestral hall murals; (4) The color expression of fresco figures in the Ming and Qing dynasties did not fit to the folk color system, indicating that artistic creativity at the period was not overly constrained by patriarchal traditions; (5) The growth in the number of female paintings indicates that throughout the age of "male supremacy over women," the human rights and value of women began to be highlighted.

According to the study, the rounded representation of the characters in the paintings linked the ornamental art of Guangdong ancestral hall murals with a back-to-human quality, asserting the value and dignity of human beings even more. To some extent, the representation and coloring of the characters in the murals is a focused reflection of ancient rites and ethical ideas.

The conclusions in sub Sect. “The evolution of sitting and lying reflected in the figure's form and rhyme - reflecting the dynamic change of etiquette from respectful to natural” are based on the interpretation of quantitative statistics on the size of the frescoes and the poses of the figures in the paintings and may be somewhat subjective. Due to the limitations of objective reasons, this paper was unable to find directly relevant evidence to support the correlation between economic cost control and the dynamics of frescoed figures. Nevertheless, the findings of this paper still provide new perspectives and aspects of reflection on this crucial aspect. The team will explore this aspect in depth in the future.