Metadata is Like Packaging: Seeing Beyond the Library Card Metaphor

From messages in a bottle to the pair of gold-anodized aluminum plaques placed on board of spacecraft Pioneer 10 and Pioneer 11 in the 1970s, how we package information is central to its visibility, durability and usefulness across different environments. And in our digital age, thе way we package information has a lot to do with metadata.

The somewhat conventional metaphor about metadata is the one of the library card. This metaphor has it that books are the data and library cards are the metadata helping us find what we need, want to know more about or even what we don’t know we were looking for. While this metaphor has been around for quite a while and still proves to be useful, we want to add one more dimension to the world of metaphors we understand metadata by.

Metadata is Like Packaging

We’ve already talked about metadata as something that enriches data with more data points that make it meaningful. It is by representing data about data that metadata augments the data with information, and makes it easier to discover, use and manage. Like everything written on a letter envelope to help the actual content – the letter – get delivered to its recipient. Or like everything written on a library index card to help the content – the book – get delivered to the reader.

Today, we can enrich that understanding with one more perspective. The one from packaging. Metadata, our CEO Atanas Kiryakov told me, in a brief conversation about Ontotext’s knowledge management solutions, is for data as packaging is for goods.

Now let’s unpack (pun unintended) this by looking at packaging and the way it is analogous to metadata.

In hindsight, the history of metadata is not very different from that of packaging. The more frequently goods were travelling and being exchanged, the more sophisticated their packaging became (see The History of Packaging). The same goes for metadata: the more intense the information environments we work in, the more complex the ways we describe the information objects in them.

Metadata communicates the meaning of data

Another close similarity between packaging and metadata is the communication of additional information. Consider food packagеs. The package helps potential buyers decide if they want to buy a product by conveying various nutritional facts, potential allergens, relevant recipes, ingredients and much more. It is the package that, by providing additional information, turns the food from just a bag with something in it into an object with information about properties, functions and potential uses. This also holds true for metadata – metadata turns unstructured, noisy data into meaningful digital objects, which makes their exchange and management easier.

Metadata facilitates the use of data

In terms of facilitating use, metadata and packaging also share similar features. Take for example spray glass cleaners. In this case, the packaging not only identifies what the product is and what it is made of, but goes even further to enhance how the cleaning liquid is used through the spray bottle technology. In the same way, metadata can serve as an integral part of the usefulness of information. For example, consider formatting tags. They help better display the product, the data they format, and in this way make it easier to consume and understand. In HTML, the formatting tags can go even further and can be used not only to define the visual appearance of the text, but also to add semantic value to it. An example is the data on a website wrapped with keywords in such a way that lets search engines consume it and understand it better so that they can classify and rank it.

Metadata protects data

The protective function of both packaging and metadata is another common feature to think about when understanding the value of metadata. Take for instance the egg carton with its multi-compartment structure housing individual eggs away from each other and preventing them from shock or damage from the outside. Similarly, many documents and formats have a checksum – “a small-sized block of data derived from another block of digital data for the purpose of detecting errors that may have been introduced during its transmission or storage”. This checksum is like a seal of integrity and shows if the received copy of the document is error free, in other words, if the eggs in the carton are still whole.

Finally, a tetra pack is far more complex than the milk we put in it and this goes even further in the cases of drug packaging. In the same way, metadata is sometimes ways more complex than the data it describes.

Doing Business By Getting Metadata Right

From the above examples, we see that metadata is as important for data as packaging is for goods.

In an enterprise knowledge management context, metadata, and especially semantic metadata, also:

• facilitates data use
• supports data reuse
• protects data

In 2021, the above three enterprise needs still require a solution.

One such solution comes from understanding metadata management as pivotal for the enterprise. Take for example publishers.

For the sake of better promotion, delivery and reuse, nowadays publishers describe and classify their content formally. It is enriched with metadata that describes it by using various fields and attributes (e.g., creation date, publication date, author, keywords, etc.), but most importantly, by references to relevant categories, subjects and entities. The latter are defined in different controlled vocabularies, taxonomies and ontologies, which represent the publisher’s preferred way of structuring and segmenting the domain of interest. Such reference data is often considered a key competitive advantage and an asset for the publisher – e.g,. BBC’s Lonclass, IET’s Subjects for engineering literature, UMLS in life sciences, FT’s giant knowledge graph of business entities, Springer Nature’s Science Graph or Getty’s thesauri for cultural heritage. Reference data enables one to structure advanced content retrieval criteria while metadata allows the actual content retrieval.

The case for organizations in other industry sectors is similar. The needs for metadata may be very different and even more diverse. Beyond controlled vocabularies, many enterprises develop and use all sorts of master and reference data, e.g., supplier lists, points of sale, product catalogues, etc. Still, these are used to provide better quality metadata, which allows employees to spend less time in finding a piece of information and to navigate company records and other databases more smoothly. And it is metadata that can serve to efficiently capture, manage and most importantly integrate enterprise data in data fabric. Serving as a reliable useful packaging for enterprise data, metadata facilitates quick and thorough sifting through research academic papers or even better writing experience for journalists and readers alike.

Epilogue: Metadata as Extensive Understanding of the Context of Data

The more information-intense our digital environments are becoming, the greater the need for well-packaged information.

Metadata, just like packaging, helps us use things more efficiently. In the case of metadata these things are digital objects. Metadata enhances the use of digital information and in terms of enterprise data, makes data easily accessible, usable and visible by stakeholders.

The more elaborate and sophisticated the metadata about data, the better data travels across information roads and cyberscapes, nicely packaged, ready to ascend the DKIW pyramid and turn into knowledge.

Do you want to package your data in a way that allows your stakeholders to know more and work smarter?

Read our white paper!