We provide an open free API for gender inference, the same one that is supporting this application. You can easily access to it through a GET request, and get quick response which is encoded in JSON.
We set no access limits, no need to sign up or pay anything, but please try not to send high-frequency requests violently. This is just a beta version for testing, and we are doing some engineering work to make it more robust.
If you may have any questions or suggestions, please feel free to contactus.
A GET request should look like this:
GET https://innovaapi.aminer.cn/tools/v1/predict/gender?name=Jie+Tang&org=Tsinghua
You should replace the "name" and "aff" parameters with full name of the user (concatenated by "+") and the affiliation information (optional).
The response is encoded in JSON, including the inference from each method along with the probability.
An example:
{
"status": 0,
"message": "success",
"data": {
"FGNL": {
"gender": "UNKNOWN",
"probability": ""
},
"WebGP": {
"gender": "male",
"probability": "96.15%"
},
"Face": {
"gender": "male",
"probability": "100.00%"
},
"Final": {
"gender": "male",
"m_proba": "98.00%",
"f_proba": "2.00%"
}
}
}
We report results from three methods: Face Recognition (Face), Facebook Generated Name List (FGNL), Web Based Gender Predictor (WebGP). The final result is generated through a voting process of all methods. Detailed information about these methods can be foundhere.
Each result consists of the inferred gender value and its probability. The field "gender" has three possible values:"male/female/UNKNOWN". The field "probability" is a real number in the range of [0, 100%]. "Final" result includes the probability to infer "male" ("m_proba") and "female" ("f_proba").