According to a recent Forrester survey, more and more businesses believe their employees spend too much time searching for the information they need: 40% today, up from 19% just five years ago. This issue is driven by a variety of issues, including a distributed workforce, siloed technology systems, and massive data growth.
But it doesn't have to be that way. Enterprise content management systems have come a long way in that same time period, and the introduction of new artificial intelligence technologies means employees can easily find and make the most of all the content their organization owns – text, audio, video, and more.
ECM has always been useful, but in the past it required a lot of effort from users. Intuitively, it's easy to understand content and categorize it according to a well-understood structure. But content management tools often required users to do things they didn't want to do, like extracting information from content and populating fields and tables to describe it. The system worked, but it required some manual effort.
AI and related technologies such as machine learning (ML) can enable content management systems to relieve users of much of the classification work. Importantly, such tools can extract relevant data even from unstructured data like PDFs, emails, and images, and accurately classify it so it's easy to find and use. Some ECM systems have intelligent document processing (IDP) capabilities that can mimic how an employee reads a document, extracts key information, and inputs it into another system for processing.
“AI enables ECM solutions to unlock valuable insights from unstructured data and maximize the value of content,” said Erica Morimoto, product marketing manager at Hyland, an intelligent content solutions provider. “Users can make more informed decisions by getting business-specific answers rather than generic answers like those from large-scale consumer language models.”
Key Features of Modern ECM
Different ECM solutions often focus on different features and use cases, but below are some of the technologies to look for in a modern ECM:
Natural Language Processing (NLP): As the name suggests, NLP uses ML to essentially “read” documents the same way a human employee would. It can perform data extraction, sentiment analysis, language detection, and document classification.
Deep Learning for Image and Video Technology: Just as NLP can “read,” deep learning technology allows ECM to look at images and videos and identify objects, text, people, activities, and more. Want to find all content that contains photos of a particular celebrity? Deep learning can help with that.
Speech-to-text: Content can take many forms, with video and audio increasing proportionately. Speech-to-text uses advanced ML algorithms to transcribe audio files into readable text that can be more easily categorised, searched and more by ECM. For example, speech-to-text can be used to transcribe customer service calls and apply sentiment analysis to determine how an agent would handle a challenging situation.
RESTful API for Image Analysis: This is another ML feature that allows ECM to classify and label images, detect embedded objects, and extract text. One use case is for insurance companies to use it to read license plates from photos of car accidents.
With such capabilities built in, an ECM platform becomes much more useful when it works across cloud environments as well as on-premise. It also makes search easier, allowing users to search and get results in natural language, just like talking to their smartphone assistant. No matter where your content is, your ECM solution can find it and make it work for you.
Learn more about Hyland's intelligent content solutions here.
