5 ms·
I'm not unsympathetic to your concerns, but the obvious retort is that if you already know which results are accurate, why bother doing analysis?
by blackbagboys 9y ago
I'm not unsympathetic to your concerns, but the obvious retort is that if you already know which results are accurate, why bother doing analysis?
- jayd16 9y agoSo you can build a model and get a computer to do the analysis...
- rspeer 9y agoPeople who make word embedding models are not doing it to find out whether men are better than women. That is not the analysis they are doing. You always have to look at your results with a critical eye to recognize when something has gone wrong with your model. You can't recognize a flaw in the model if you believe everything the model says.
- blackbagboys 9y agoI'd be pretty amazed if it was even possible to construct a model that would answer that question. In the article's example, you're constructing a model to determine how a movie might be reviewed. The data indicates that movie reviewers may be reviewing movies with male leads more positively than movies with female leads. There are a lot of possible reasons for this other than "men are better than women", so why would you simply reject the result out of hand?