Conclusion

This thesis has presented a schema for evaluating and improving heritage interpretation in the form of the five ideals of heritage interpretation, surveyed, evaluated and suggested improvements for current Korean cultural heritage interpretation resources, processes, and content, and demonstrated the possibilities of data-based heritage interpretation as a solution to the weaknesses of current Korean cultural heritage interpretation through the design and implementation of an ontology suitable to a labeled property graph. It contributes to the fields of heritage interpretation, digital humanities, and Korean studies in various respects.

First, the ideals of heritage interpretation presented in this thesis, though founded in prior scholarship, consider a broader conception of heritage interpretation than existing definitions and principles, and therefore can be used as evaluative yardsticks for a wider variety of interpretive resources. They are not specific prescriptions of how heritage interpretation must be done, and therefore can be realized in a variety of ways which take into consideration the unique circumstances (i.e. the nature of the interpretive information, financial and human resources available, current technological infrastructure, etc.) of each institution and individual implementing them as evaluative criteria. Therefore, these ideals will be useful for continued innovation of heritage interpretation resources and processes, not only in Korea, but anywhere in the world.

Second, this thesis surveyed and evaluated current Korean cultural heritage interpretive resources, processes and content to a greater extent than prior research. Prior research had investigated grammatical errors, inconsistencies, in the content and translations of interpretive texts, surveyed some online interpretive resources, or made prescriptive and broad suggestions for the content of interpretive texts. This thesis went beyond such research – analyzing with greater depth a wider variety of available interpretive resources, breaking down by heritage type the specific elements and structures commonly found in interpretive texts, as well as explaining the process by which interpretive texts are created and translated by local governments. The thesis also presented an abbreviated translation of the guidelines of heritage interpretation as provided by the CHA.

This research will be useful to the CHA and local governments, which may be able to more clearly understand the nature of the interpretation work they are attempting to do, as well as the fundamental changes in structure and mindset which need to be made to solve current shortcomings and make future improvements. It may also put greater pressure on the organization to make such changes if the public and scholars abroad are made aware of the many problems with current interpretive resources. In addition, though but a byproduct of the process of evaluating the current status of Korean cultural heritage interpretations, the overview of currently available interpretive resources and the breakdown of content in interpretive texts will aid those interested in traditional Korean culture and history – be they tourists, content creators, students, or scholars – in gaining a better understanding of the kind of resources available. As mentioned in the thesis, these resources (both online and offline) are unfortunately separated by institution, and not well linked to, advertised, or explained, which are speedbumps in the discovery and utilization of such potentially useful resources.

Third, the development of an ontology and examples of its implementation demonstrate how interpretive information can be turned into data yet still convey its full context. In this researcher’s anecdotal experience, there is a general perception that somehow the complexity of the historical and cultural content contained in interpretive resources could not possibly be conveyed through data, nor could such data be used as the basis of meaningful interpretive resources. As will be discussed below, there are still many improvements which need to be made to the ontology, as well as further research into the way to use algorithms and interfaces to turn data into interpretive resources, in order to create an ontology – and by extension, data-based interpretive resources – which convey the context with the highest fidelity possible. However, even the crude, preliminary examples shown in this thesis demonstrate that complex concepts, events, objects, and their relationships to one another can be stored as data and transformed into various presentation forms which convey nearly the same nuance seen in interpretive texts, while facilitating a level of personalization of content and display which is simply not possible with interpretive texts written by humans. Previous ontologies as introduced in this thesis had only developed ontologies or data models with the objective to describe the interpretive information and use it as data, not reutilize it in interpretive resources. In this respect, this thesis makes unique contributions to the investigation of how to design a database which does not exist merely as a database, but can be reutilized into a variety of resources. Thus, this ontology opens doors for future Korean studies research, education, content creation, etc. resources, while also more broadly aiding other digital humanities scholars in non-Korea, non-heritage interpretation areas in the development of ontologies which can be reutilized in resource generation.

However, there are, of course, various limitations of this research. First, although the researcher thoroughly reviewed existing heritage interpretation definitions and principles to the best of her ability distill the five ideals of interpretation presented in this thesis, there may be other ideals, or other ways of conceptualizing such qualities interpretation should strive to embody. Further research on this front may help to develop better criteria with which to review and improve interpretive resources. In particular, the conceptualization of heritage and heritage interpretation is limited to Western sources, while there is no consideration given to the influence of Japanese or industrialization-era conceptions of heritage in Korea today. In the future, further investigation should also be made regarding Joseon-era ideas of cultural heritage and historical sites.

Second, while this thesis surveyed various interpretive resources, the process by which interpretive texts in particular are created and translated, and the content of select on-site interpretive texts, this review could have been more extensive in regard to the scope of heritages types and resource types reviewed. As mentioned, there was particular difficulty in getting verifiable information on the interpretive resource creation process, and this could be an area of further investigation. There is also a need to investigate the status of interpretive resources for a wider variety of heritages, including museum artifacts and intangible heritages.

Third, the ontology presented in this thesis has various weaknesses which needs to be improved before it can be realistically implemented. Most urgently, there needs to be further research on how to represent more complex relationships via a graph database, which are not simple “A has type B” relationships, but are “A visited B sometime between C1 and C2 with D because of E which led to F” and so on. The ontology presented in this thesis does successfully utilize relationship properties and node IDs to convey these various prepositional modifications to actions in ways which allow the node properties to be reutilized in query results. However, this information is currently very difficult to query due to the fact that the particular prepositional properties and the labels of their contents stored in each relationship vary from case to case; there is no single straightforward query to cover all relationships. Furthermore, there is no way to include a reference or meta information (i.e. who added/edited it, when it was added/edited) for each of the various prepositional properties in a relationship; Currently, only one reference and one meta-information can be stored for each relationship. These prepositional properties could be stored each as separate relationships, but that raises the question of how to link together those various relationships to show that they are related to the same central relationship. Furthermore, this adds to the number of relationships in the database, which may avoidably add to the size of the database. Therefore, further testing of the ontology to uncover the most data-efficient and easily query-able model must be undertaken. In addition, further research of interpretive resources for a broader range of heritages should be undertaken to better grasp the nature of the contextual elements and their relationships to one another. Following this, testing with test users regarding what kind of information they would want to filter would also help refine labels and relationship types to maximize real-world usability.

Furthermore, a key potential benefit of such a data-based heritage interpretation is the incorporation of the various contextual elements and relations into the larger Semantic Web. However, since the Semantic Web is based on RDF/OWL, research to see if this ontology can be represented in RDF, not just as a labeled property graph, is needed. Furthermore, identification and reutilization of equivalent properties and relationships in other widely used data models and ontologies, such as EDM and LIDO, will further facilitate the incorporation of such a database’s contents into the larger world of cultural heritage data.

Fourth, some features of data-based heritage interpretation and the ontology presented in this thesis which had been mentioned in the thesis do not have examples in Section VII due to limited space in an already lengthy thesis. In future research, additional examples should be included to demonstrate the full range of ways in which these approaches specifically address the current weaknesses of Korean cultural heritage interpretations and more fully realize the five ideals of interpretation.

In addition to addressing the weaknesses in this thesis, there are various steps that must be taken looking forward in order to bring the ideas presented in this thesis into reality. One of the main downsides of data-based heritage interpretation is that it requires large-scale, long-term oversight by an institution which has the financial resources to facilitate it – unlike interpretive texts which can be created in a relatively short time and then forgotten about (see Kim et al 2016, 189). An institution in charge of such a database would need the authority to establish a standard process which could be followed by all the other organizations that would contribute to and utilize the database. This likely means that such an institution would need to be a government institution, though possibly academic or private. As mentioned throughout this thesis, the South Korean government’s rotational bureaucratic system does not lend itself to successfully seeing-through such long-term projects which require skilled human resources. Therefore, figuring out what kind of institution could successfully spearhead such a project and what relationship said institution has with the South Korean government will need to be considered in-depth before actually beginning such an undertaking. Otherwise, the dream of data-based heritage interpretation may be doomed to meet an early death – not due to any fundamental problem with the concept of data-based heritage interpretation itself, but due to a mishandling of its implementation.

While the financial and authoritative capabilities for the facilitation of data-based interpretation lie in the hands of such institutions, without preliminary research such as that presented in this thesis, such innovations in heritage interpretation would not be developed, precisely because of the bureaucratic divisions which lead to a lack of innovation in heritage interpretation as discussed in Sections III.2. and IV. However, this researcher believes that a mere ontology or database model will not persuade such institutions to see the benefits of data-based interpretation. To persuade such institutions, the development of a functioning and visually pleasing interface that allows non-tech-savvy individuals to input, filter, search, analyze and transform the data into interpretive resources of various forms (visualizations, text, timelines, tables, diagrams and more) will be crucial. The development of such interfaces also naturally requires research of potential users, their motivations, and the way they actually use such interfaces.

Another issue which will need to be settled will deal with standardizing the process by which information is added to the database. This issue is related to questions raised by Staiff (2016) in the context of the democratization of heritage:

While government or academic institutions have the obligation and financial/bureaucratic power to develop better interpretive resources for the public, it is the very people in charge of such organizations – bureaucrats and academics – who may have the most resistance to the idea of data-based heritage interpretation. They may be such an approach as a direct challenge to “the authorized voices of heritage specialists and the highly regulated and controlled canons of masterworks subject to protection regimes,” and be “concerned about not ‘letting go,’…more interested in protecting their professional practices,” (Staiff 2016, loc. 2960; 2892). These parties, even if not oppositional due to a desire to protect their authority and control messages, may also be simply unfamiliar with digital technology, rendering them unable to imagine the benefits of digital and data-based methods. Therefore, navigating this kind of opposition and ignorance may become a key task in the realization of data-based heritage interpretation.

By extension, such individuals may have strong opinions regarding the qualifications of those who can manipulate and access the data in the database, as well as processes for fact-checking information and proofing translations. Despite the fact that current heritage interpretations have abysmal quality control, such a database model where information is centralized may lead some to argue that only academics and professionals should be able to enter data. There is also a risk of possible censorship of certain facts which run counter to the desired narrative of the government or private owners of heritages. In an effort to empower non-experts (such as civil officials and local citizens, both of Korea and other countries) to input diverse data that is meaningful to them while also ensuring the accuracy and quality of said data, the ontology presented in this thesis includes features to identify the creators and editors of information in the database, as well as ways to include references and inform users when information is not verified but merely presumed, so that users are empowered to make their own judgements on veracity. Furthermore, whether data is immediately added to the database (and fixed later if there are problems), or whether data undergoes a basic fact-checking and proofreading protocol first before being officially added to the database, etc., also would need to be decided upon.

However, any difficulties in navigating such details and convincing overprotective authorities will be worthwhile in the long run. Data-based heritage interpretation will provide new pathways to meaningful and creative engagement with Korean traditional culture for an active, diverse, and global public. As alluded to in Section II.2, data-based heritage interpretation is not meant to be a complete replacement for “analog” pathways to interpretive information, such as personal introductions to Korean cultural heritages and experiencing them in “real life” – whether that be at a heritage site or a museum. An algorithm-based suggestion for a heritage site may not have the same emotional impact as the recommendation of a dear friend, and the experience of a heritage via a virtual reality medium cannot (yet) compare to the visceral experience of seeing and experiencing both tangible and intangible cultural heritage in person. Instead, we must remember that the data-based perspective toward heritage interpretation is not a direct substitute for traditional (i.e. analog, one-directional, narrative) interpretation methods, but rather meant to 1) address the otherwise un-addressable shortcomings of such methods (such as access for people who cannot visit in person, physical limitations of content length, depth, and personalization, multi-functional usability of interpretive information, etc., as mentioned throughout this thesis) and 2) simultaneously act as a platform not only for heritage interpretation in its traditional sense of “educating the public,” but also for academic research, student-led research and exploration, and creative content sourcing and creation.

Data-based heritage interpretation is aimed at minimizing redundancy while maximizing automation where possible, which can then free up human resources to focus on those elements of interpretation which need a human touch - such as methodology research, fine-tuning of content and processes, and person-to-person connections. In this way, the database (and the algorithms and interfaces via which it is manifested into content) serves not only to facilitate more personalized and contextualized content which can be used for multiple purposes and can adapt to changing technology, but more fundamentally to improve efficiency for more sustainable and innovative heritage interpretation. Indeed, by minimizing redundancy and automating content creation, a data-based approach to heritage interpretation may, in the long run, ultimately free up more time and energy to spend in improving the analog heritage interpretation experience than is currently available.

Though the upfront investment to develop a database and the interfaces to go along with it may be significant, the database can be enriched far into the future, ultimately becoming key resource for scholarly research in addition to its use as a tool of heritage interpretation, while new interfaces with which to access and present the data can continue to be developed as technological trends change. In the future, such a database could, and should, be expanded to include not just heritage of South Korea, but that in North Korea, China, and other sites of Korean diaspora. Combined with virtual or augmented reality technology, such a database could allow us to step back into time and venture into ‘physical’ places we cannot access in real life for practical or political reasons. Through the connections stored within the database itself, we can render and make accessible the past as a part of the present rather than relegating it to the confines of archives and museums, allowing us to better understand how it matters to our contemporary, everyday life. By minimizing the gap between the superficial consumption of “traditional Korean” culture and the complex, ever-negotiated context of Korean heritages, data-based heritage interpretation not only prompts audiences to reflect upon how Korean tradition and heritage has a place in their lives today, but brings non-academics closer to the world of Korean studies. It is by bridging these kinds of divides in ways that are relevant to the present and considerate of the future that Korean tradition and Korean studies will continue to be meaningful to people long into the future.