top of page
  • Writer's pictureIlia Zelenkin

How to Extract Data from Any Document using Bitskout

Updated: Sep 10, 2023


In this guide, we will learn how to use Bitskout's new feature to extract data from any document. This feature allows you to easily extract information from different types of forms by providing examples.


Video Instruction






Step 1.

Log in to your Bitskout account and click Create Plugin.



Step 2.

The next step is to click on the Extract button and then choose "From File":




Step 3.

Now we need to load an example. Bitskout needs examples from you to learn what you'd like to extract. We will use SEC filling reports as an example.



Step 4.

Add examples to guide the data extraction process. Once you've loaded the file, write the fields that you'd like to extract. You need to give the field a name and the value should come from the loaded examples. See below:




Step 5

You can add more examples to improve the accuracy of the data extraction. Let's add another form with a different layout.


Once the file is loaded, you'll need to add the required values from this example:




Step 6

Once you're done with examples, press Create and the plugin will be created. Next, depending on what you'd like to do, choose a tool where you want to use the plugin.



Step 7.

Verify the accuracy of the data extraction by testing it on various forms. For example, try using an "Apple 10-Q form":


Conclusion

Now your plugin is ready to be used. This way you can extract data from the documents with just a few examples.


We recommend adding 2-3 examples that have varying layouts. This way Bitskout will understand the variety of the documents that it'll have to analyze the information.


You can load documents in any language - the most important part is to add clear values from that example to extract.





14 views0 comments
bottom of page