Massive Time Saver! Auto-generating a Code-Frame using Codeit’s Text Analysis

08 July 2019

Massive Time Saver! Auto-generating a Code-Frame using Codeit’s Text Analysis
Codeit’s unique integrated text and sentiment analysis capability can easily and quickly auto-generate a new code-frame. The newly generated code-frame categories will be based on the key themes and sentiment found in the verbatim content.

Once the code-frame has been generated, Codeit’s easy to use code-frame editing tool allows the user to arrange, edit and merge the initial suggested categories into a final version for coding and reporting.

image.png

Aside from the time-saving involved in letting Codeit create the code-frame from the verbatims that need to be tagged, another important benefit of this approach is that a large percentage of the verbatims can then be automatically coded, speeding up the process even more. As the autogenerated code-frame is created from the actual key phrases within the verbatim comments, the coding is highly productive and subsequent use of the code frame on new data will provide consistent results.

And lastly, once the initial coding is completed by a combination of the automated text-matching and human coding staff, the categorized examples can train the machine learning coding layer to further augment overall productivity. The advantage of this approach is that the resulting code frame can use all the “smart coding” layers built-in to Codeit, meaning that any subsequent waves of data will be automatically, and consistently, coded using the previous coding and machine learning.

image-20190708095053-1.png

Whilst these simple automated steps dramatically improve project setup and coding productivity, importantly, it’s not at the expense of quality or accuracy. Codeit’s automation is always applied under the control of human coding staff who have the final say in choosing to apply any or all the Codeit smart-coding layers - in any order. Codeit highlights every phrase where a specific code is suggested by any of the smart coding layers - the results can be swiftly reviewed and are completely transparent to anyone making recommendations based on the coding.

12.jpg

A further layer of quality security is provided by the new Machine Learning Confidence Score assigned to each suggested coding result.

Auto-code frame generation can dramatically speed up development and implementation of coding and tagging projects, reducing turn-round times and costs, without diminishing quality.

Please contact us here to schedule a demonstration of this powerful Codeit capability using your own verbatim content.



Back to Blog