During the summer of 2019, BYU Harold B. Lee library teamed up with ZONTAL, Inc. to see if...
How to make your data reusable
Data is the most valuable commodity of the digital age; therefore, you need to treat it well. Read this post to learn how to make you data reusable, and how ZONTAL offers the functionality to transform your business.
Data has been dubbed “the new oil”, in an apt comparison with the commodity that ruled last century’s economy. Just like oil, data needs to be refined before it is truly useful, in a process that transforms raw information into refined assets with practical applications. This reusable data fuels analytics tools and artificial intelligence algorithms, leading to faster product creation and giving organizations unique strategic insights. But what does refined data look like, and how do you get it to that state?
What reusable data looks like
Broadly speaking, reusable data has a number of characteristics that make it easy to work with. Firstly, it is structured in a consistent way, so you know what to expect and can easily use it in visualizations or statistical analyses. Secondly, it is self-explanatory, meaning that you can understand what is going on when you look at it, without having to ask the creators for explanations. Thirdly, it is easy to find, so that you can quickly locate it and use it even years after it has been produced. These attributes are captured by the FAIR guiding principles, which explain – among other points – what makes data reusable. To learn more about this subject, please refer to our whitepaper on FAIR data. Ever since these guidelines were published in 2016, organizations that handle scientific data have been gradually updating their management strategies to keep up with the current demands of the industry. By and large, the consensus is that the answer to achieving reusable data lies in adopting an effective ecosystem of tools.
Making your data reusable
A core component of this ecosystem is a data repository that provides the infrastructure for keeping data standardized and easy to access. Once you have chosen a suitable repository system, you can make your data reusable by working through the following steps.
Step 1: Identify data silos
The first step is to identify where your data is currently stored. Different departments in your organization may have unique data storage locations, so it is important to communicate with all teams to find out where these are located. The goal is to get rid of these isolated data storages – commonly called data silos – and instead create a unified data layer, but depending on the existing architecture that might not be immediately possible. In those cases, the content from these silos can be replicated in the data layer, giving the organization time to restructure applications to use this new source of data.
Step 2: Transfer data to the new location
Once you have found all of the data you would like to refine, it is time to transfer it from the existing silos to the new storage location. This depends on the needs of the organization, with some selecting cloud storage, others opting for local storage, and a third group combining elements from these two solutions. When choosing between these options, remember to take into account issues such as data safety and scalability. As you consider the right solution for you, keep in mind that ZONTAL supports cloud, on-premise, or hybrid deployment options. Once this step is complete, all your data will be accessible under a single system, and it becomes possible to remove the now redundant data silos.
Step 3: Harmonize data
Next, you should update the structure of the data to follow a consistent pattern that enables reuse. Ideally, this harmonization process will preserve different types of information – e.g. experimental results and instrument observations –, while also storing metadata that provides additional context to the records. Ultimately, the restructuring should produce self-reporting data assets, meaning that the information is organized and annotated in such a way that users can interpret it without any additional help. Another benefit of harmonization is that it makes large amounts of data interoperable between systems, allowing others to directly consume the information produced by applications.
Step 4: Ensure data integrity
Finally, it is time to verify the integrity of your data to ensure that it has the highest possible quality. At a fundamental level, this process can be thought of as a data clean-up, where you make sure records have the correct value types and ranges. However, a more sophisticated approach consists of cross-checking the data against an existing standard. In the context of data integrity, this means ensuring that the metadata used to describe experiments and observations employs a vocabulary – known as reference and master data – that is consistent throughout the organization. If you have opted for using ZONTAL, you can seamlessly integrate with external systems, such as Accurids, for checking against master and reference data. This step makes your data reusable, as it is now clearly annotated with descriptions.
Where to go from here
After reviewing these steps, the task of making data reusable may appear daunting, but do not fret: Our ZONTAL team is up to the challenge and ready to support your efforts. You may be wondering “fair enough, but where do I go from here?” While the initial transformation described here provides some benefits, a holistic approach that includes systems as well as processes will lead to lasting benefits. We can help you improve your effectiveness and give you the upper hand in data reusability. More crucially, we will guide you in developing robust data management principles to prepare you for an increasingly digital future.