Evans, James and Pedro Aceves. 2016. “Machine Translation: Mining Text for Social Theory”. Annual Review of Sociology 42:21-50. DOI: 10.1146/annurev-soc-081715-074206
Sentiment analysis has good results because it has broad application, so many people work on developing dictionaries and exploring new methods. I am curious about the generalization of supervised algorithms for other (less common) topics. Training algorithms seems to be expensive. For specific projects, is it feasible to train algorithms which may not be directly useful for other studies?
Sentiment analysis has good results because it has broad application, so many people work on developing dictionaries and exploring new methods. I am curious about the generalization of supervised algorithms for other (less common) topics. Training algorithms seems to be expensive. For specific projects, is it feasible to train algorithms which may not be directly useful for other studies?