Gender Inference Methods
Depending on the type of data you obtain for analysis you can use one of the three following methods for discovering gender:
Images only
Image-based methods rely on visual data for gender inference. The data can be provided as web-links pointing to images or as a batch of image files. In our ”images only” section you can find information on how to use these methods, which output the inference models provide, and what limitations need to be considered.
Names only
Most methods for Inferring gender from names are based upon large name dictionaries or datasets. Some algorithms also take country of origin or time period into account. Check out our "Names only" page to learn about further details and corresponding datasets for these models.
Mixed
If the data you want to use for gender inferences includes both images and names, check out the model we describe on our mixed page. Mixed methods have a good chance to outperform methods that rely on names only or images only as they can use more information for their predictions.
Hint: if you already have names you can apply this algorithm to extract Google images from it. This way, you can use a mixed method for more accurate and precise results.