Smart Metadata and Smart Extractor are two intelligence services that M-Files provides, both of which can bring powerful data automation and extraction tools to your document management system. However, upon initial review, it can be hard to tell the difference between the two services.
The purpose of Smart Metadata is to analyze vault contents and pick up patterns of historical metadata entry, thereby producing metadata suggestions for new or existing objects/documents within the vault. Under the hood, Smart Metadata relies on M-Files Knowledge Graph which maps out the data relationships and patterns, thereby allowing it to learn and suggest properties such as:
These suggestions for data entry and extraction are very similar to what you'll see from the M-Files Smart Extractor, which consists of two services - Matcher, and Text Analytics. The key difference here is that while the Smart Extractor will suggest any relevant data within the document (data which it has been configured to find), Smart Metadata will pay attention to where in the document the data was found. For example, if you always add a date value from the second page of the document and not the first, Smart Metadata will know which date to suggest.
In short, where M-Files Extractor compares the contents of the document to the contents of the vault, Smart Metadata compares the contents of the document to the contents of the vault AND the historical patterns of data entry from the users.
After reviewing the difference between Smart Metadata and Smart Extractor, you might immediately think that Smart Metadata is the way to go; after all, it sounds like Smart Metadata provides the same data automation and extraction capabilities as Smart Extractor, but with additional intelligence added. This is close to the truth, however, Smart Metadata varies in its success depending on the contents of your document management system.
Depending on the robustness of data within your vault, you might not receive optimal suggestions from Smart Metadata. Smart Metadata is an intelligence service that requires training, and thus many existing examples of documents with correct metadata entries are required before Smart Metadata will be able to provide data extraction and entry suggestions.
In this case, if your data falls into the below two categories, Smart Extractor may not be the best choice for data automation:
In these cases, the M-Files Smart Extractor may be a better choice for your system. However, if your vault has a lot of accurate textual data which can be utilized in training, Smart Metadata can be a powerful tool for use within your vault.
Interested in learning more? Read about additional intelligence tools in our other insights, or reach out to TEAM IM at www.teamim.com.