Developing an Evaluation Framework for Chinese ICH Short Videos in International Communication: A Multi-Evidence Design ()
1. Introduction
The convergence of globalization, digital media, and mobile viewing has made short-video platforms important channels for presenting intangible cultural heritage (ICH) across cultural and linguistic boundaries. Visual demonstration, compressed storytelling, and networked circulation allow craft procedures, performance traditions, local knowledge, and community memory to reach audiences who may have little prior knowledge of China. Accounts such as Grandpa Amu have shown how process-centered narratives, limited verbal explanation, and recognizable human relationships can support international circulation without reducing ICH to a conventional promotional message [1]. Short videos, therefore, offer both an opportunity for wider cultural access and a challenge of representing living heritage accurately in platform-driven environments.
Greater visibility does not necessarily produce deeper understanding. ICH videos may attract views while leaving historical context, community ownership, or the social meaning of a practice unclear. Platform conventions can also encourage repetitive aesthetics, compressed explanations, and entertainment-oriented editing. These tendencies may increase immediate attention but weaken cultural specificity. Cross-cultural communication adds another layer of difficulty because subtitles, images, sound, narrative assumptions, and audience knowledge interact in ways that cannot be captured by a single popularity metric [2] [3]. The evaluation of ICH short videos must therefore address not only reach but also cultural accuracy, interpretive accessibility, emotional response, participation, and safeguarding implications.
Existing research has generated useful measures, but these measures were developed for different purposes. Some studies evaluated digital dissemination systems or cultural value [4]-[6], whereas others examined cross-cultural effects, visual narrative, user needs, branding, or platform promotion [7]-[10]. Their units of analysis also varied among videos, accounts, platforms, comments, users, and institutions. Consequently, a direct combination of all available indicators would create conceptual overlap and measurement problems. A useful framework must specify what each indicator measures, where its data come from, and whether it represents a communication condition, a content feature, an audience response, or a longer-term outcome. Against this research background, this paper constructs an operational evaluation system, so that creators, cultural institutions, and researchers can judge cross-cultural communication performance of ICH short videos with transparent, multi-faceted evidence instead of merely counting online view volume.
2. Literature Review
2.1. Communication Effects and Audience Reception
Research on user needs has moved beyond simple engagement counts by examining why viewers attend to and share ICH content. Wang et al. [9] identified information, emotional, and interaction needs and showed that content-quality satisfaction was necessary for sharing in their case, while emotional resonance, social interaction, and topic selection appeared in several configurational paths. This approach is useful because it recognizes that sharing can arise from different combinations of motivations. However, its evidence came from derivative videos associated with one cultural work and from Chinese-language platform comments. The identified needs are therefore analytically valuable but cannot be transferred to overseas audiences without additional validation.
Yu [10] distinguished contact, understanding, and identification as progressively deeper levels of cross-cultural communication. Li et al. [8] similarly organized YouTube responses around cognition, emotion, and behavior. These frameworks share a sequential logic: exposure enables interpretation, interpretation can support affective response, and affect may precede behavioral intention. Their strength lies in separating visibility from deeper outcomes. Their limitation is that platform traces do not always demonstrate psychological change. A complete view, a positive comment, or a purchase-related statement may suggest attention or intention, but these traces do not by themselves establish comprehension, identification, or actual behavior. Evaluation systems should therefore distinguish publicly observable behavior from audience-reported outcomes.
Value-centered studies broaden evaluation beyond communication efficiency. Xu et al. [4] classified cultural, social, and economic value in a large corpus of Bilibili comments and found particularly strong performance for cultural value perception. The computational approach offered scale and replicability, but its categories remained dependent on the language, platform, and training data used. Yan et al. [5] used qualitative interviews to identify 30 indicators related to digital dissemination, while Yan et al. [6] combined stakeholder interviews, a Delphi process, and the Analytic Hierarchy Process to construct a weighted system for Shu brocade. The two Yan studies offered unusually explicit indicator-development procedures, although their focus on a single traditional technique limited direct generalization to other ICH forms and to international audiences.
Cao [7] examined musical ICH promotion through a questionnaire-based TikTok study and linked platform features, content strategies, participation, and perceived outcomes. The study demonstrated the value of measuring user perceptions rather than inferring them entirely from platform counts. At the same time, platform-specific constructs such as special effects or challenge mechanisms can change quickly and may not be comparable across TikTok, YouTube, Bilibili, and Douyin. Stable evaluation dimensions should therefore describe observable communicative functions, while platform-specific features should be treated as adaptable operational indicators.
2.2. Content Quality, Narrative, and Branding
Studies of international communication frequently emphasize emotional accessibility, process-centered storytelling, and localized explanation. Sun and Gu [1] showed how the Grandpa Amu account combined craftsmanship, everyday relationships, and visually intelligible action. Li [11] examined speed, visual form, and plot construction from a visual-culture perspective, while Jiang [12] discussed narrative strategies in internet-distributed ICH documentary films. These studies explain how narrative form shapes attention and interpretation, but they mainly provide analytical categories rather than validated scales. Their concepts overlap substantially: narrative clarity, emotional resonance, visual coherence, authenticity, and audience accessibility appear under different labels. This overlap supports synthesis, but it also requires explicit rules for merging indicators.
Zhao and Bai [13] developed a short-video evaluation scale around narrative, rhetorical, linguistic, and technical abilities. Although the scale addressed foreign-language teaching rather than ICH, its separation of message design from technical execution offered transferable categories. Wang and Yang [14] proposed a mechanism linking cultural essence, public accessibility, and value realization for traditional-dance videos. Wu [15] analyzed material, imagined, and lived dialogic spaces in popular ICH videos. The former provides a clear process model, whereas the latter offers interpretive depth. Neither alone resolves how abstract qualities should be scored consistently across coders.
Brand-oriented research examined content positioning, symbol systems, content matrices, and media integration [16]. Ouyang and Wang [17] organized communication paths around actors, spaces, and content, and Guo [2] connected content quality, channels, and communicative settings to cross-cultural barriers. These models identify organizational conditions that influence performance, but they do not measure the same object as a video-level quality scale. Branding, institutional support, and account continuity are therefore best treated as contextual dimensions rather than direct evidence that a particular video produced understanding or identification.
2.3. Cross-Cultural Adaptation and Cultural Representation
Multimodal research emphasizes that subtitles, spoken language, images, music, gesture, and editing jointly shape meaning. Wu [3] argued that coordination among verbal and nonverbal modes can improve the translation of ICH promotional videos. Zhang [18] likewise proposed moving from isolated cultural symbols toward shared values and audience-oriented interpretation. These perspectives are especially relevant to ICH because many practices rely on tacit knowledge and context. Nevertheless, multimodal richness can increase rather than reduce interpretive difficulty when visual symbols are unexplained or when subtitles translate words without explaining cultural relations.
Cross-cultural adaptation creates a tension between accessibility and cultural specificity. DeSoucey et al. [19] described authenticity as a socially organized and transnationally negotiated claim rather than a fixed property. This insight cautions against treating official certification, traditional costume, or an unchanged performance setting as automatic evidence of authenticity. The UNESCO [20] Convention also defines ICH as living practices recognized by communities, groups, and individuals. Accordingly, evaluation should consider cultural accuracy, community participation, consent, and continuity while allowing responsible innovation. Adaptation is not necessarily cultural loss, and preservation is not equivalent to static reproduction.
2.4. Reflections on Previous Research
The first gap concerns conceptual integration. Existing studies repeatedly identify content quality, cultural representation, narrative clarity, technical production, emotional resonance, interaction, and value perception, but they organize these concepts in different ways. Some frameworks treat them as production inputs, whereas others treat them as communication effects or cultural outcomes. Without a shared structure, overlapping indicators may be counted twice, and variables at different analytical levels may be combined inappropriately.
The second gap concerns cross-cultural specificity. Studies based on Chinese platforms provide detailed evidence about domestic users and local media practices, while international-platform studies offer evidence about YouTube or TikTok circulation. However, overseas audiences are often treated as a single group, and cultural background, language, and prior knowledge are rarely incorporated into the measurement design. The available international literature is also too limited to support a broad claim that foreign models are mature but unsuitable for the Chinese ICH. A more defensible conclusion is that the evidence remains uneven across cultural settings and requires targeted validation.
The third gap concerns operationalization. Qualities such as spirit conveyance, aesthetic integration, and narrative originality are legitimate evaluative concepts, but they are not automatically quantitative indicators. They require observable criteria and anchored rating scales. Conversely, platform metrics such as views and comments are observable but may not be comparable across platforms or may be unavailable to researchers. A useful framework must therefore combine objective metrics with transparently defined qualitative ratings and must mark unavailable data as missing rather than treating it as zero.
The fourth gap concerns the boundary between communication performance and heritage outcomes. Reach, comprehension, cultural respect, community cohesion, sales, tourism interest, and long-term safeguarding operate over different time periods and require different evidence. They should be related in a process model rather than collapsed into one unqualified effectiveness score. The present study addressed these gaps by first constructing a broad preliminary framework.
3. Research Design
3.1. Research Questions
The study addressed two research questions:
RQ1: Which indicators were most relevant to the production quality and communication effectiveness of Chinese ICH short videos in international settings?
RQ2: How could these indicators be organized into a framework that remains culturally sensitive, measurable, and usable across different platform environments?
3.2. Data Collection
A focused literature corpus was assembled to support indicator construction. The purpose was not to claim exhaustive coverage of every publication on digital heritage. Four practical selection criteria were used. Each study addressed ICH, short-video communication, or a closely transferable short-video evaluation problem. It also presented explicit dimensions, indicators, or measurement procedures, described its methods clearly enough for comparison, and was available in full text. This approach was appropriate because the study required usable indicator structures rather than a prevalence estimate of the entire field.
Fourteen studies were selected to provide diversity in research object and method. Four were English-language international journal articles, and 10 were Chinese-language journal articles. The corpus included qualitative interviews, Delphi and AHP procedures, computational comment analysis, content coding, questionnaire surveys, case studies, and conceptual analyses. It also covered YouTube, TikTok, Bilibili, Douyin, and Kuaishou. Table 1 summarizes the analytical contribution of each source. The focused design increased comparability, but it also limited representativeness; this limitation was addressed explicitly in the discussion. The literature-based indicator work was complemented by a preliminary field investigation in a heritage community, reported in Section 3.4, which checked the framework’s relevance against first-hand transmission conditions.
Table 1. Indicator frameworks included in the focused literature corpus.
Source |
Framework |
Main Contribution |
Method and Data |
Yan et al. (2025a) |
Qualitative indicator exploration |
30 indicators emphasizing authenticity, cultural integrity, skill transmission, and stakeholder participation |
Semi-structured interviews; NVivo; grounded-theory coding |
Yan et al. (2025b) |
Weighted evaluation system |
Technology, culture, communication, transmission, and policy dimensions |
Stakeholder interviews; modified Delphi; AHP; Shu brocade case |
Xu et al. (2024) |
Value-perception classification |
Cultural, social, and economic value |
23,130 Bilibili comments; BiGRU-Attention; topic analysis |
Li et al. (2024) |
Cognition-affect-behavior model |
Awareness, image, emotion, learning, purchase, sharing, and travel intentions |
YouTube comments; keyword, sentiment, and co-occurrence analysis |
Wu, S. (2025) |
Dialogic-space framework |
Material, imagined, and lived spaces; labor narrative; emotional community |
Popular Bilibili videos; observation, case analysis, and textual analysis |
Ouyang & Wang (2025) |
Communication-path framework |
Communication actors, spaces, and content |
Douyin and Kuaishou case analysis |
Guo (2025) |
Foundation-barrier-strategy framework |
Content quality, narrative, interaction, channels, and cultural barriers |
Conceptual analysis and cases |
Li & Fan (2023) |
Brand-communication framework |
Positioning, symbol system, content matrix, and media integration |
Short-video account case analysis |
Yu (2026) |
Cross-cultural effect hierarchy |
Contact, understanding, and identification effects |
Video coding, qualitative modeling, and regression-based evidence |
Li (2026) |
Visual-culture
analysis |
Narrative speed, visual form, plot construction, and authenticity |
Conceptual analysis and cases |
Wang & Yang (2026) |
Three-part communication mechanism |
Cultural essence, public interface, and value realization |
Comparative Bilibili cases; word-frequency and topic analysis |
Wang et al. (2025) |
User needs and
sharing model |
Information, emotional, and interaction needs; sharing configurations |
BERT classification; sentiment analysis; fsQCA |
Zhao & Bai (2024) |
Short-video
evaluation scale |
Narrative, rhetorical, linguistic, and technical abilities |
Scale development for Chinese
story short videos |
Cao (2025) |
Platform-promotion framework |
Platform features, participation, content strategies, and perceived outcomes |
TikTok questionnaire; regression,
t tests, and ANOVA |
Note: Yan et al. [5] refer to the Sustainability qualitative-interview article; Yan et al. [6] refer to the Heritage Science Delphi and AHP article. Years and study descriptions were checked against the source files.
3.3. Data Analysis
The 14 sources were decomposed into their stated dimensions, categories, variables, and operational measures. Terms were retained in their original form during the first coding pass so that conceptual distinctions were not lost prematurely. This process produced 125 candidate indicators. The coding procedure used open, axial, and selective coding as a practical organizing sequence rather than as a claim that a new grounded theory had been generated from primary field data [21].
Two researchers independently compared the candidate indicators by meaning, analytical level, and proposed data source. Indicators were merged when they referred to the same underlying construct, differed mainly in wording, and could share an operational definition without losing substantive content. Disagreements were discussed with a third researcher. For example, cultural authenticity, integrity of cultural connotation, authenticity in labor narratives, and cultural essence were grouped under cultural authenticity and essence presentation. Sharing intention, recommendation intention, and sharing behavior were retained as related but distinct indicators because intention and observed behavior require different data. Platform-specific features were grouped under technology only when they described observable content or interaction affordances. The procedure reduced the pool from 125 to 72 indicators. During operationalization, indicators that depend on inaccessible platform analytics were replaced with observable or clearly qualified alternatives: for example, algorithm matching was removed in favor of coder-rated platform fit or within-platform public reach, completion rate was replaced by average viewing time where the platform reports it or otherwise marked as missing, and unverified overseas-viewing proportions were replaced by within-platform geographic breakdowns or audience surveys. Measures that remained unavailable were marked as missing rather than treated as zero.
The 72 indicators were organized into 10 preliminary dimensions and three functional layers. Policy, support, and branding were treated as contextual conditions. Technology, culture, communication, transmission, and cross-cultural communication are described in the content and communication processes. User needs, behavior, value perception, and overall evaluation represented the audience and outcome domains. The layer structure clarified causal order but did not imply that all indicators could be added to a single score (See Table 2).
Table 2. Preliminary 10-dimensional structure derived from the focused literature corpus.
Dimension |
Layer |
Categories |
Representative Indicators |
Policy and support |
Context |
Government and legal support; platform and institutional support |
Policies, intellectual-property protection, cultural-institution collaboration |
Branding |
Context |
Positioning; symbol system; content matrix; media integration |
Account positioning, visual identity, thematic consistency, multi-platform coordination |
Technology |
Process |
Production quality; multimodal integration; platform adaptation; stability and cost |
Image and audio quality, editing, metadata clarity, format adaptation |
Culture |
Process |
Authenticity; diversity; regional context; craftsmanship |
Cultural accuracy, community voice, local specificity, safeguarding orientation |
Communication |
Process |
Contact; understanding; identification; interaction; dialogic space |
Public reach, clarity, resonance, replies,
co-creation |
Transmission |
Process |
Community participation; knowledge continuity; transmission challenges |
Practitioner involvement, procedural demonstration, skill-loss risk |
Cross-cultural communication |
Process |
Cognitive; emotional; behavioral; cultural barriers |
Contextual accessibility, linguistic appropriateness, respect,
behavioral intention |
User needs and behavior |
Outcome |
Information; emotional; interaction; sharing; participation motivation |
Knowledge, enjoyment, belonging, recommendation, learning motivation |
Value perception |
Outcome |
Cultural; social; economic |
Cultural value, community benefit, product or tourism interest |
Overall evaluation |
Outcome |
User satisfaction; communication impact |
Content satisfaction, viewing satisfaction, longer-term visibility |
Note: The table summarizes the preliminary organization of 72 indicators.
3.4. Field Investigation
To check the framework’s relevance against real heritage conditions, the literature-based synthesis was complemented by a preliminary field investigation. In September 2025, the research team spent two days in Shangyang Village, Yiwu City, Zhejiang Province, where two local ICH practices, the Zhudeng (beaded lantern) and the string-operated lion dance, were studied through semi-structured interviews, site observation, and hands-on inspection of craft tools and performance equipment. Interviewees included a village committee representative, a craftsman who had participated in restoring the Zhudeng lanterns, and the inheritor of the string-operated lion dance; the visit was planned with the village committee, and interviews were conducted with the interviewees’ consent.
Three findings were directly relevant to the framework. First, community knowledge is rich but weakly documented: the craftsman described the complete lantern-making process, from bamboo framing and beading into pictorial patterns to painted decoration, including the wire-clamping technique, and its use in local festivals, yet much of this knowledge exists only in practice and memory. Second, transmission is precarious: the inheritor reported aging practitioners, limited younger successors, and uneven records of performance conventions and of the line-and-pulley control system that operates the lion. Third, communities are willing to collaborate in external communication: both the craftsman and the inheritor expressed openness to cooperation, indicating that practitioner involvement and community voice can be observed directly rather than inferred.
These findings map onto the preliminary framework in two ways. They support the Culture and Transmission dimensions, where cultural accuracy, regional context, practitioner involvement, knowledge continuity, and transmission challenges reflect conditions that communities themselves report. They also indicate that sign interpretability and procedural demonstration are the points at which current video content most often fails, a pattern examined in the six-sample trial assessment in Section 4. A complementary online-fieldwork case study of bamboo-weaving videos reported similar bottlenecks, including inadequate multilingual translation, weak content localization, and superficial audience engagement [22].
4. Trial Assessment
4.1. Sample Profile
To examine the applicability of the preliminary framework, six representative overseas-distributed ICH short videos were selected for a trial assessment. The trial assessment serves three purposes: to test whether the 72 indicators in the 10-dimensional framework can be observed and coded in real videos; to show how content features, audience responses, and platform conditions are distinguished within the framework; and to identify the indicators that current practice most often leaves unaddressed.
The six samples were purposively selected to span craft types (velvet flowers, mortise-and-tenon joinery, bamboo weaving, Yi lacquerware, Ru celadon, and Chengdu lacquer art), platforms (YouTube and Douyin), video lengths, account types, and levels of international visibility. The trial assessment does not validate the framework statistically; it provides an applied check of relevance and measurability before expert consultation (See Table 3).
Table 3. Profile of the six trial-assessment samples.
No. |
CraftType |
Title |
Platform |
Account |
Duration |
Views |
Likes |
Comments |
Date |
S1 |
Ronghua (绒花) |
Blue Rosewood Poem Silk Flower Hairpin |
YouTube |
@lanseichina |
29 s |
9.89 M |
200 K |
631 |
2025-07-11 |
S2 |
Mortise & Tenon (榫卯) |
阿木爷爷榫卯打造一座木工桥,全程无钉子,高手在民间【阿木爷爷Grandpa Amu】 |
YouTube |
@GrandpaAmu |
13 m 54 s |
59 M |
620 K |
28 K |
2019-12-20 |
S3 |
Bamboo Weaving (竹编) |
#Intangible Heritage Basic Style Doesn’t Match Basic Style #Bamboo Weaving Crafts #Intangible Heritage |
YouTube |
(made by AI) @aikiduniya-i1c |
12 s |
176 M |
810 K |
3187 |
2025-09-26 |
S4 |
Yi Lacquerware (彝族漆器) |
漆器椅子 |
Douyin |
@彝三彩 |
34 s |
- |
528 K |
13 K |
2019-04-27 |
S5 |
Ru Celadon (汝瓷) |
The Color of Sky After Rain: Ru Celadon雨过天青云破处–汝瓷 |
YouTube |
@山白Shanbai |
10 m 04 s |
6.07 M |
64 K |
2703 |
2025-03-27 |
S6 |
Chengdu Lacquer Art (成都漆艺) |
奶奶的衣柜坏了,给她翻新了一下。My grandma’s wardrobe was broken, so I gave it a makeover. | Liziqi Channel |
YouTube |
@李子柒Liziqi |
14 m |
36.43 M |
950 K |
66 K |
2024-11-12 |
Note. Data collected as of August 2026. Views and engagement metrics are subject to change.
4.2. Content Feature Analysis
1) S1—Ronghua Hairpin (YouTube, @lanseichina)
The video employs a concise, step-by-step visual narrative: individual velvet components are fabricated and progressively assembled into a complete hairpin. Minimal narration relies on close-up handcrafted shots. The 29-second duration compresses the full making process, prioritizing visual transformation over verbal explanation. Hashtags (#velvetflowers #ronghua #silkflowers) facilitate discoverability but provide no cultural context.
2) S2—Mortise & Tenon (YouTube, @GrandpaAmu)
At 13 minutes 54 seconds, this is a long sample. It documents the complete joinery process—from raw timber to a wooden bridge—without nails or glue. The outdoor workshop setting and near-absence of dialogue create an authentic, meditative viewing experience. However, the video lacks a textual or verbal explanation of the historical significance of mortise-and-tenon joinery in Chinese architectural tradition.
3) S3—Bamboo Pot-Scrubber (YouTube, @aikiduniya-i1c)
A 12-second fragment showing an elderly man riding an AI-generated bamboo-woven electric vehicle with a child on the back, surrounded by cheerful onlookers. The video carries a striking novel visual appeal with no on-screen text or voice-over. The real-world constraints of bamboo craftsmanship and the genuine cultural connotations of bamboo-weaving heritage are entirely omitted.
4) S4—Yi Lacquerware (Douyin, @彝三彩)
This 34-second video focuses on the painting process: traditional mineral pigments in red, yellow, and black are applied to a wooden base. The strong color contrast produces high visual impact. Yet the symbolic meanings of the three-color scheme in Yi ethnic cosmology remain unarticulated—the video presents aesthetic form without interpretive depth.
5) S5—Ru Celadon (YouTube, @山白Shanbai)
A 10-minute-4-second fragment documenting Ru porcelain production, including glaze-material preparation, wheel-throwing, trimming, and glazing procedures. The video features meditative slow-motion cinematography and subtle ambient sound. The critical high-temperature reduction-firing phase with nail-spur support, as well as Ru ware’s status as the pre-eminent Song-dynasty imperial kiln and its historical discontinuation, are largely unelaborated.
6) S6—Chengdu Lacquer Art (YouTube, @李子柒Liziqi)
A 14-minute sequence documenting lacquer harvesting, body-repairing, repeated lacquer coating, hidden-relief carving, gilding, and polishing of a refurbished wooden wardrobe. The video features serene cinematic visuals and descriptive English subtitles, weaving a warm intergenerational-affection narrative. While hands-on craft procedures are well documented, the long-standing history of Chengdu lacquer art and the cultural connotations of the qilin motif are not elaborated on in detail.
4.3. Commonalities and Limitations across Samples
Common strengths identified:
1) Visual-centric communication. All six samples prioritize close-up footage of handcraft processes, minimizing verbal narration. This strategy lowers cross-linguistic comprehension barriers and relies on universal visual grammar.
2) Process compression. Short-form editing condenses lengthy craft procedures into digestible segments, emphasizing the satisfying “raw material to finished object” transformation.
3) Authentic setting. Most samples feature real workshop or domestic environments rather than staged studio settings, conveying authenticity and artisanal credibility.
Common limitations identified:
1) Decontextualization of cultural history. None of the six samples systematically explains the historical origins, dynastic associations, or evolutionary trajectories of the featured crafts. Cultural knowledge is subordinated to visual spectacle.
2) Absence of inheritor narratives: The craftspeople remain largely anonymous—their names, generational lineage, community roles, and personal stories are not introduced. The human dimension of ICH transmission is invisible.
3) Symbolic opacity. Decorative motifs, color schemes, and material choices that carry specific cultural or ethnic meanings (e.g., Yi lacquerware’s three-color cosmology) are presented as pure aesthetics without semiotic explanation.
4) Fragmented process representation. Several samples show only isolated stages of production, omitting critical steps such as firing, dyeing, or curing, which prevents audiences from grasping the craft’s full technical complexity.
These limitations correspond to specific indicators in the preliminary framework. Decontextualized cultural history indicates weak evidence for cultural accuracy and regional context (Culture dimension); the absence of inheritor narratives indicates missing practitioner involvement and community voice (Transmission dimension); symbolic opacity points to limited sign interpretability and contextual accessibility (Cross-cultural communication dimension); and fragmented process representation signals incomplete procedural demonstration and craftsmanship indicators (Culture and Technology dimensions). The common strengths map onto accessibility-related indicators: visual-centric communication supports contextual accessibility, and authentic settings support authenticity and cultural accuracy. In framework terms, the six samples perform well on visual accessibility and production quality but poorly on cultural accuracy, sign interpretability, and practitioner participation.
4.4. Analytical Implications
These six samples collectively reveal a dominant pattern in overseas Chinese ICH short video dissemination: craft as visual spectacle, culture as silent backdrop. The videos successfully attract cross-cultural attention through universal visual appeal, but they risk reducing living intangible heritage to decontextualized aesthetic objects. The trial assessment shows that the preliminary framework captures this pattern: the samples score well on visual accessibility and production quality, whereas cultural accuracy, sign interpretability, and practitioner involvement remain largely unaddressed. The tension between accessibility and cultural depth is therefore not the core object of the present study; it is the reason why evaluation must treat accessibility and cultural depth as separate indicators rather than collapsing them into a single popularity score. Section 5 uses these findings to specify how the framework should be operationalized.
5. Discussion
5.1. Conceptual Contribution
The proposed framework contributed a layered view of ICH short-video communication. Two contextual dimensions, policy and support, and branding, described institutional and account-level conditions that shape communication without proving its effects. Five process dimensions, technology, culture, communication, transmission, and cross-cultural communication, described message characteristics and content, and communication processes. Three outcome dimensions, user needs and behavior, value perception, and overall evaluation, captured audience responses that require direct measurement. This structure reduced the risk of treating every favorable condition as evidence of communication success.
The framework also treated cultural accuracy and cross-cultural accessibility as related but nonidentical goals. A culturally detailed video may remain difficult for unfamiliar audiences, whereas a highly accessible video may oversimplify the heritage practice. Evaluation should therefore examine how videos negotiate this tension. Community-informed assessment provides evidence for cultural integrity, while target-audience testing provides evidence for accessibility and response.
5.2. Methodological Implications
The framework should be used as a set of linked modules rather than as an automatic total score. Video-level content ratings can be combined within validated subscales. Public metrics should be analyzed within platforms because view and engagement systems differ. Audience scales require psychometric testing and, for international comparison, translation procedures and measurement-equivalence analysis. Contextual and long-term outcome indicators require institutional or stakeholder evidence. If a composite index is eventually created, its weighting and aggregation rules should be justified through expert evidence and sensitivity analysis.
5.3. Practical Implications
For creators, the framework provides a pre-publication checklist centered on cultural accuracy, practitioner participation, narrative clarity, translation, and platform-specific presentation. For cultural institutions, it distinguishes documentation and safeguarding goals from short-term visibility. For researchers, it supplies a common vocabulary and proposed measurement sources. For platform or account managers, it encourages transparent reporting intervals and discourages claims based on private metrics that cannot be independently reviewed.
Even with this evaluation framework in hand, source-material constraints may still persist for real-world ICH short-video productions. Where original video material carries inherent weaknesses such as thin contextual information, superficial practitioner participation, or oversimplified narrative, supplementary fieldwork investigation can remedy indicator-framework limitations. Through field interviews with heritage inheritors and local community stakeholders, researchers can supplement missing contextual evidence for the evaluation system. For external communication practice facing defective source material, promotion strategies can shift focus: highlight observable craft action sequences, add layered multilingual annotation, adopt shared-value framing instead of dense cultural background exposition, and match the release rhythm to overseas platform users’ habits to offset deficiencies in the original content. The field investigation reported in Section 3.4 illustrates this logic: the lantern craftsman’s process knowledge and the lion-dance inheritor’s transmission concerns supplied contextual evidence for cultural accuracy, procedural demonstration, and transmission-challenge indicators that none of the sampled videos currently provides.
5.4. Limitations and Future Research
The focused corpus did not constitute an exhaustive systematic review and remained weighted toward studies of Chinese platforms and Chinese ICH. The proposed thresholds, indicators, and scoring procedures, therefore, require expert revision. Future work should report the full indicator-decision trail, recruit experts with international and community-based experience, conduct multilingual audience testing, and examine whether the measurement structure is stable across platforms, ICH categories, and cultural groups. Ethical review, informed consent, privacy protection, intellectual-property considerations, and community benefit should be documented before primary data are collected.
6. Conclusion
This study reorganized fragmented research on ICH digital dissemination, short-video production, cross-cultural communication, and audience response into a transparent framework-development design. A focused set of 14 studies generated 125 candidate indicators, which were consolidated through literature-based synthesis into a 72-indicator preliminary framework organized into 10 dimensions and three functional layers. A trial assessment of six representative overseas short videos showed that the framework can be applied to real content, and a field investigation in a heritage community supplied first-hand evidence of the transmission and participation conditions that the framework is designed to capture. The preliminary framework avoided reliance on inaccessible measures such as algorithm matching, completion rate, and unverified overseas-viewing proportions, and combined public metrics, anchored qualitative ratings, audience measures, and contextual evidence without assuming that they formed a single validated score. Expert consultation and cross-platform testing are required before the framework can be treated as a validated instrument.
Author Contributions
Conceptualization: Huang, H.X., Lv, Q.R., and Lu, J.Y.; Methodology: Huang, H.X.; Software: Huang, H.X.; Validation: Huang, H.X., Lv, Q.R., and Lu, J.Y.; Formal analysis: Huang, H.X.; Investigation: Huang, H.X., Lv, Q.R., and Lu, J.Y.; Resources: Huang, H.X., Lv, Q.R., and Lu, J.Y.; Data curation: Huang, H.X.; Writing—original draft preparation: Huang, H.X., Lv, Q.R., and Lu, J.Y.; Writing—review and editing: Huang, H.X. and Lu, J.Y.; Visualization: Huang, H.X.; Supervision: Huang, H.X. and Lv, Q.R.; Project administration: Huang, H.X. All authors have read and agreed to the published version of the manuscript.
Funding
This article is supported by the 2026 East China University of Science and Technology Student Innovation and Entrepreneurship Training Program project “AI-Supported Research on the International Communication of Chinese Intangible Cultural Heritage: Innovative Short Video Production and Effect Evaluation” and the 2026 East China University of Science and Technology Undergraduate Education and Teaching Reform Research Project “Interdisciplinary Course Construction” category project “Simultaneous Interpretation”.