KNIME is a free and open-source data analytics, reporting and integration platform. KNIME integrates various components for machine learning and data mining through its modular data pipelining concept. A majority of the functionality is provided by the modules of the platform, while other functionality is provided by KNIME extensions. KNIME offers a complete platform for end to end data science,
I am spending hours on my job manually adjusting data entries due to inconsistent spelling and shortcuts, so I built a tool to automate this process using large-language models.
The idea is to have a tool that takes in CSV exports and:
> Corrects for inconsistencies in spelling (Coop vs co-op)
> Harmonizes shortcuts (Limited vs Ltd.)
> Corrects for spelling mistakes (serbices vs services)
This is how the tool works:
You can upload a CSV file and specify which row you want to extract and harmonize.
The model automatically consolidates data by combining similar-looking phrases.
You can edit the proposed phrase names or further consolidate entries if there are some groups the model has missed.
In the end, you can download your CSV file again and push it to the database