July 7, 2017

Transparent Systems: The Role of AI and Machine Learning in Content Management

4 minute read

Manual content management takes up too much time to be productive. As we said in our last post in this series, employees in traditional/paper-based offices spend six hours a week on average searching for paper documents and eight hours per week creating reports. With productivity at such a premium, this time investment is not sustainable.

In the approaching future, artificial intelligence (AI) and machine learning will leverage vast amounts of business 

information to intelligently define and enforce governance rules with minimal need for human input.

Why AI?

A great place to start on this topic is a quote from a recent article on CMSwire from our own Mike Alsup:

"Legal, HR and Business Departments will collectively decide what content needs to be governed from a privacy, access and retention perspective and will transparently enforce these rules at a content/tag level. Containers and templates will automatically (and transparently) apply governance policy."

The main takeaway from this quote is that by asking individual departments to define what data they deem to be important, the organization can implement have the proper automated workflows and retention policies. This allows the records team to maintain control of critical data without negatively impacting the processes of the end users (or overworking your records staff).

As we have discussed in prior blog posts, this transparent and automatic governance ensures accuracy and simplicity while mitigating compliance issues. A traditional ECM, however, relies on users manually classifying documents. This is not only time-consuming, it can also lead to crucial errors and data between departments being siloed.

Where does machine learning fit in?

Machine learning is the process by which software learns from frequent and common actions that users take, and turns them into automated workflows. The benefits here are obvious:

  • More efficient processes
  • End users are no longer required to manually perform these tasks
  • The software will work 24/7 and always be updated
  • Document-intensive businesses will be positively affected

Machine learning can then be applied to organize and analyze these large batches of information much more effectively. It will also enable easier searches for the relevant content. Machine learning will organize the information going forward and add structure to previously unstructured content. Depending on the level of machine learning implemented, it can also recommend the appropriate tags and metadata.

Practical application of these technologies

Now that we’ve discussed the scope and benefits of these solutions, let’s discuss how they can benefit an organization in the real world.

As unstructured content continues to grow at an increasing pace, developing automation to govern it becomes more and more important. Even the most dedicated information governance team will be unable to effectively handle the demands required from this volume of information.

Big data (and the mining of it for business use) continues to show results for companies. However, this big data is useless if it is not being properly governed and managed. For a business to effectively use big data insights, they must first have complete visibility of all that data, as well as the ability to trust the integrity of that data.

Information governance expert Jeffrey Ritter makes an important note about what data is being fed into a company's big data analytics technology: "The engines are performing crunching on the numbers, but they can only crunch numbers that filter in correctly -- the data has to qualify, it has to meet the rules that are useful for the engine to compare to find the patterns, and if the data does not qualify, it is embargoed -- it is quarantined or it is rejected."

By setting up proper information governance and the rules around it, the data can actually be used to drive new revenue by uncovering new business intelligence instead of becoming redundant, obsolete, or trivial (ROT.)

Finally, every business is concerned about reducing cost across the board. By using AI and machine effectively, employee productivity can be increased, which in turn, reduces cost. Instead of salaried employees spending their time on administrative tasks, they can focus on their core competencies. Additionally, by automatically removing data with no business value, companies can reduce their overall data volume, and thus their storage costs.

Receive News Updates As Soon As They Happen