Constructing effective search queries

When searching within Enigma Public, it’s important to understand how information is organized and how search operates. This will help you find the information you’re seeking. If you haven’t already done so, review the articles How information is organized and Searching for information before continuing.

When you search Enigma Public, you are actually performing three separate searches:

  • The first search looks for matches within the collection metadata
  • The second search looks for matches within the dataset metadata
  • The third search looks for matches within the data records in each dataset

To illustrate how this works, consider a search for facebook h-1b visas, which yields no matching results.

Enigma Public has a collection called “H-1B Visa Applications,” which contains multiple datasets – one for each year.

Each dataset includes multiple visa applications from Facebook. Below is the 2017 dataset filtered using the keyword “Facebook.”

So why did our search yield zero results?

Examining the screen above you can see that:

The parent collection is called “H-1B Visa Applications”

The dataset description includes the words “H-1B” and “visa”

The dataset title also includes the words “H-1B” and “Visa”

Multiple rows include the words “H-1B” and “FACEBOOK”

Importantly, though, no single category includes all of the search terms, so there are no matching results. Just as importantly, no category includes the word “visas” (only “visa”).

So how should you locate information on H-1B visa applications by Facebook within Enigma Public? Generally, it’s best to identify the kind of information you are seeking first (in this case, information about H-1B visa applications). Then either filter the search results in the Data Viewer based on the entity of interest (in this case, Facebook), or examine the resulting datasets to understand how best to construct an effective search query. In this example, examining the datasets suggests that a search for facebook h-1b would yield useful results.

When searching, keep in mind that a search for “visas” is different from a search for “visa.” If you’re not finding the information you want, consider using the fuzzy matching operator, ~1, which signifies “within one character of,” so visas~1 would match both the singular and plural forms.