This is a working prototype that I developed for GSA's AI Community of Practice in order to explore data preparation techniques for NLP and create a strategic approach to data preparation aligned with the user's data attributes.
Acting as an initial guide, it aids in identifying approaches suitable for a range of NLP challenges, promoting the exploration and further adaptation to meet the specific requirements of an NLP project. Please note that as a "proof of concept" it is not yet fully functional, but is configured to generate differentiated results.
The user can specify data characteristics by adjusting the sliders, then click the "submit" button at the end of the page to receive tailored data preparation recommendations. A script is used to generate a quantitative score based on the settings of individual sliders, whose sum is used to generate the NLP prep recommendation.