How Verdictr Works

How Data is Extracted

Image

Verdictr is incredibly easy to use. After navigating to the keyword search page  go to the search box and start searching! However, you might want to know how we extract our data. While we can’t go into full detail, we can explain the basic process. This process starts with us scraping decision texts from the European Commission's websites. After scraping these texts, we use highly intelligent and fined tuned large language models to analyze the data. After the data has been analyzed, we clean the data and upload it to our database.

Searching Practices

Image

Before you start searching, you might want to know how the regex works. When you input a word, you will only get exact matches back. For example, if you enter "Bi ke" into the search bar, you will not get matches with the text "Bike" back, only matches with the text "Bi ke." That is because the searches are space sensitive. However, the searches are not case sensitive; both "Bike" and "bike" will yield the same results. You might run into special characters within you market decisons. If you find yourself in this spot, note that special characters are included in the search function. There are more do’s and don’ts than the ones discussed above. For more information, visit our FAQs page.

Searchable Data

Image

We have extracted 5 main data values from market decision; the case number, link to case, link to decision text, topic, and market decision text. The case number—the number assigned by the European Commission to identify cases—is the case number of the case where the market decision is located. The link to case is the link to the case's page on the European Commission's website. The link to decision text is the link to a PDF that contains the market decision. The topic is the general idea of what the market decision. For example, the decision text "A computer is an electronic device for storing and processing data, typically in binary form, according to instructions given to it in a variable program" might have a topic of "Computers." The market decision text is the text extracted from the decision text PDF. The market decision text has undergone minor filtering, where some special characters and misplaced spaces were removed. You will not get results if you add the data values in your search. For example, if you search for "Case Number: M.4696" you will not get any results, however, if you search for "M.4696" you will get results for any market decisions extracted from case M.4696.