Web Analytics Made Easy - Statcounter

USE CASE

Autoclassification and autoappraisal

Metadata plays a major role in information management. It’s used to classify, describe, find and discover, protect, appraise and dispose of records. It’s also essential for effective decision-making, research, analytics, and information sharing.

You need fit-for-purpose metadata if you’re going to successfully manage the ever-growing stockpiles of information. But you can’t rely on staff to manually classify documents. It’s time-consuming, difficult and, if we’re being honest, boring. You need a smarter way to assign metadata that does the job properly.

With the right knowledge and tools, you can stop working for your information and get your information working for you.

Your staff can breathe a collective sigh of relief, while your organisation benefits from:

 Increased productivity

 Improved decision-making

 Reduced risk

 Lower operating costs

It’s about working smarter, not harder.

How Synercon can help

Synercon have substantial experience in deploying autoclassification to deliver records management outcomes, chiefly in the SharePoint environment. We can help you generate high quality fit-for-purpose metadata which will improve findability, automate appraisal and disposal, enable migration and apply information protection.

Tools

We work with a suite of tools to deliver our automated solution.

a.k.a. software by Synercon simplifies the task of developing taxonomies, thesauri and ontologies for deployment into an autoclassification engine. With a.k.a. we codify contemporary records management instruments, business classification schemes and disposal schedules into machine readable formats.

Synercon SuperSaurus enables us to fast track the production of taxonomies, thesauri and ontologies.

DiscoveryOne Content Enrichment engine for processing autoclassification and autoappraisal of records across multiple environments and tagging records within the SharePoint.

Percipio RM software by Synercon for processing the disposal of records within the SharePoint environment.

Techniques

We’ve developed considerable knowhow on how to deploy autoclassification.

 Our autoclassification methodology is based on a faceted classification model from which we autogenerate fit-for-purpose metadata, which enhances findability and drives records management processes.

We employ a rules-based approach to autoclassification which is upfront, observable and measurable.

We transform disposal schedules and security classification schemes into data models (ontologies) which drive automated appraisal, to tag records with security and disposal metadata.

We have developed simple to use techniques for analysing and measuring the precision and coverage of tagging.

We advocate an iterative, measured approach to deployment, enabling benefits to the realised throughout the project.

We integrate with the SharePoint Term Store and Managed Metadata to truly enable the SharePoint search experience.

Training

Successful autoclassification is much more that working with the technology.  It requires multiple competencies to cover all of the bases. Our training program is designed to address all aspects of autoclassification from end to end, to build self sufficient teams.

Quick links

Being shovel ready for Microsoft 365
Intelligent automation of recordkeeping – and approach that really works
Barriers to autoclassification for information governance and how to overcome them

FAQ

Frequently asked questions

What is autoclassification?

Autoclassification is a methodology that that classifies content by analyzing its text and other properties, and through the application of rules and/or algorithms,  automatically assigns labels (or tags) to the document based on a pre-defined class.

Autoclassification can use rule based models, which produce pre-defined outcomes that are based on rules coded by humans; or machine learning models, where the system defines its own set of rules that are based on data outputs. ​

Each system has its own advantages and disadvantages. Machine learning models are probabilistic and use statistical rules rather than the deterministic approach of a rule-based AI model. Machine learning systems require significantly more data than rule-based models to achieve accurate results. Rule based models are recommended when there is a high chance of error and when fast results are required. 

What are the benefits of autoclassification?

Productivity

Business users just don’t have time to classify the high volumes of content that they manage. Even an expert would take 2-3 years to process what an autoclassification engine can process in one night.  Autoclassification enables business users to get on with their day jobs.

Coverage

Once you start the process of automated crawling and classification – you can ensure that no documents are missed  or overlooked.

Consistency

Autoclassification delivers consistent metadata, because the rules that are established for auto-classification are applied consistently across every piece of content. It is this consistency that makes search work, enabling the refiners and filters in SharePoint and other search engines.

Accuracy

Autoclassification is a knowledge based process. The more knowledge your feed into your engine, the better it will classify.  Fine-tuning your taxonomies and associated rules leads to high levels of precision, in many cases 85-95% for specific taxonomies.

    Was this page helpful?

    Insights

    Related posts

    Got a question?

    Send us a message and we'll do our best to help.

    Pin It on Pinterest