Metadata: What it is and how it works

The definition of metadata sounds simple — it’s a set of data about other data. In other words, it provides information and context for the data. Metadata makes finding, managing, using, and understanding the data it describes easier. For example, if you took a picture with your smartphone, the metadata of the digital image would include:

• The data type — the file’s name, for example.
• How the data was created — the type of camera it was taken with.
• When the data was created — the date and time the picture was taken.
• Where the data was created — the location the picture was taken at.
• How big the data is — the size of the file.
• The author of the data — the phone the picture was taken with.

Each camera or app might capture additional metadata and slightly different types of metadata. However, the metadata of a picture will tell you information about the picture but will not show you the image itself. The same goes for the metadata of digital assets — it will provide information about the asset but not its actual content.

Metadata summarizes basic information about other data and is broadly used in information management and automated information processing. You generate metadata each time you create, modify, or delete a document, file, or any other information asset. You can either embed metadata in a data asset or associate it with the asset. For example, you can embed metadata in digital assets using a special format or store it in the document’s properties.

Metadata describes data assets to help users understand their purpose and relevance, outlines the organization and relationships of different components within a data set, facilitates administrative data management, ensures long-term accessibility and integrity of data, helps to search for the data and discover it, and provides technical details for software and devices to understand and display each data element correctly.

Metadata language is standardized so humans and computers can understand it. This standardized language allows for better interoperability and data integration across applications and computer systems.

Metadata also plays a vital role in the creation of web pages because it conveys essential information for users and search engines. Web developers use HTML tags to embed metadata in the web page code. These metadata tags provide descriptions, titles, and keywords summarizing the web page’s content and purpose. Meta tags appear in search results provided by search engines, so they are used to optimize search engines and make data discovery easier for the user.

Users also generate metadata on web pages by leaving comments, liking, sharing content on social media, and even searching for information on a web page by typing in queries. Businesses use this user-generated metadata for SEO, content personalization, market research, and product improvement.

Metadata existed long before the digital era. In libraries, catalog cards provide information — metadata — on books, such as the titles, authors, and topics. But the first mentions of the word “metadata” date back to the late 1960s.

In the early computer era (1960s-1980s), file systems already included metadata like file names, sizes, and creation dates. In the 1990s, developers introduced the Dublin Core Metadata Element Set, a simple standard for describing digital resources. First developed to aid in indexing physical library card catalogs, the Dublin Core metadata standard is now the main standard for information discovery on the Web.

Around the early 2000s, due to the rise of the internet and web development, tech specialists started using meta tags in HTML to indicate web page descriptions and keywords. Metadata gained prominence in digital photography, music, and video. Developers started embedding metadata in digital assets.

Today, metadata plays a vital role in data management, content discovery on the internet, search engine optimization, and data analysis. Metadata is also crucial in data governance, usage, quality assurance, and compliance. It is also used extensively in developing artificial intelligence (AI) models because it provides context and details about the training data.

There are twelve types of metadata. The main three types are descriptive, structural, and administrative. Still, it can be further classified into smaller types according to the specific role that it plays in data management:

Descriptive metadata is bibliographic information identifying specific data by describing its content, context, and characteristics. This type of metadata includes the document’s title, creator’s name, data type, creation date, number of volumes, keywords, etc. This type of metadata may also include a summary of the resource’s content.

Structural metadata is like the table of contents in a book — it provides information about the physical organization, arrangement, and relationships between data files and resources within systems. For example, the structural metadata of a video might include information on the sequence of different parts of the video and where the ads play.

Administrative metadata is information about the resource’s origin, the data’s creator or owner, access permissions, and other copyright and usage rights and policies.

Technical metadata provides information about the technical attributes of a digital resource, such as file size, format, resolution, encoding, data source, metadata schema, and other technical data.

Process metadata offers information about the results of various operations and workflows in a data warehouse. Process metadata contains details on different actions, tools, and steps taken to generate, modify, and manage statistical data. This type of metadata helps to analyze data and understand its quality and reproducibility.

Preservation metadata includes long-term data preservation information, such as file migration history and digital signatures. This type of metadata ensures that digital resources remain accessible, authentic, and understandable over time.

Legal metadata provides information on licensing, copyright, ownership, usage rights, and other legal aspects of digital resources.

Usage metadata records information about the usage of data sets. It might include the number of views a digital asset has received, who accessed it, when, how long users interacted with it, etc. Businesses use this type of metadata to assess their assets’ popularity and analyze customer behavior to improve their products and services.

Quality metadata is information about the quality level of data. It measures data quality, accuracy, currency, reliability, and completeness of the data. It details on dataset statuses, freshness, tests run, and test success.

Statistical metadata includes information about data collection methods, sampling techniques, and data accuracy. For example, this type of metadata ensures transparency and reliability of statistical reports and surveys published by government agencies.

Reference metadata provides information about the semantics and relationships of data elements in larger datasets or databases. This metadata type helps users to understand the meaning and interconnectedness of data elements, which helps to interpret and analyze the data.

Collaboration metadata contains information on interactions, contributions, and communication around the data. It includes data-related comments, chat transcripts, tags, bookmarks, and issue tickets, which helps to collaborate and work more effectively.

You don’t have to work in data management to come across examples of metadata — you use it in your everyday life on your smartphone and laptop. Imagine you want to download a new app on your phone. You will probably read through other users’ comments before installing the app. These user comments and ratings are examples of descriptive metadata.

File names help you locate relevant digital resources on your computer or smartphone, while metadata in tags lets you find the content you need online. And you also generate metadata when you work with digital documents: when you add or delete some of the content, make changes to the heading, font, or layout, leave comments, or send it over email.

You also generate metadata when posting pictures or videos on social media sites as well as by adding descriptions, captions, or hashtags to your posts.

Metadata is vital for communicating information about digital content. It makes finding, using, and managing data much easier by providing a standard mechanism and language. Finding the files you need or any other content online or in a database would be challenging if there were no metadata tags that describe the content. Metadata also helps track all changes and data exchange.

Metadata plays a vital role in data security. By specifying who can access the information, metadata ensures that sensitive data is protected and only authorized users can view or modify it.

Careful metadata management and protection is crucial for your data privacy and security. Cybercriminals might use various techniques to place malware within the metadata of legitimate files and other digital resources. Therefore, you should ensure that no unauthorized persons can access your files and their metadata. Here are some tips on how to protect your metadata:

• Use a VPN. A virtual private network (VPN) encrypts your online traffic while it travels the internet. Your internet service provider will be unable to see what you do online or track your online activities that might generate more metadata, like comments on social media sites or search queries. A VPN also hides your IP address, which can be traced from metadata from websites you have visited.
• Implement access control. Restrict the access to files and their metadata to authorized users only. Allow different access levels (view only, edit, share, etc.) to different users.
• Authenticate users. Allow only authenticated users to have access to digital assets. Use strong, unique passwords for accounts and implement multi-factor authentication for extra security.
• Use encryption. Encrypt files and data to ensure they are safe even if unauthorized access occurs.
• Mask and anonymize sensitive metadata. Mask sensitive metadata or anonymize it to remove personally identifiable information when sharing the files with individuals who are unauthorized or do not require full access to the data.