CrushDetector

免费增值
通过 meimessaging | 已更新 4 months ago | Text Analysis

CrushDetector 概览

关注者:2
资源:
产品网站 使用条款
API 创建者:
avatar
meimessaging
meimessaging
给 API 打分:
登录并给 API 打分

README

Predicting Relationship Outcomes

Our Crush Detector API is for dating app developers that want to create the next Tinder or Bumble, just better. By implementing our API, you can offer your users AI to predict what the conversation will likely turn into. You can simply send the messages (anonymously of course) to our AI via the API, which can use the words to determine whether this relationship would end up being platonic or romantic. For example, it will come back with “Sympathetic friend” or Passionate lover”.

How Does It work?

To train the model to give the probability score, we have hundreds of thousands of conversations where users have labeled as romantic or not romantic relationships. Because of the amount of data and labels we have, we’ve been able to build neural networks/AI models that can accurately predict the type of relationship from the text messages that are fed into it, without asking the user. We have similar models for personality that can build personality profiles.

**Predicting Relationship Outcomes
**
Our Crush Detector API is for dating app developers that want to create the next Tinder or Bumble, just better. By implementing our API, you can offer your users AI to predict what the conversation will likely turn into. You can simply send the messages (anonymously of course) to our AI via the API, which can use the words to determine whether this relationship would end up being platonic or romantic. For example, it will come back with “Sympathetic friend” or Passionate lover”.

**How Does It work?
**
To train the model to give the probability score, we have hundreds of thousands of conversations where users have labeled as romantic or not romantic relationships. Because of the amount of data and labels we have, we’ve been able to build neural networks/AI models that can accurately predict the type of relationship from the text messages that are fed into it, without asking the user. We have similar models for personality that can build personality profiles.

打分:5 - 投票:1