Hello,
Can someone give me an idea on how deep learning applied in the analysis of open-ended survey questions?
My intuition is that applying deep learning models for analyzing the vast majority of survey questions is infeasible since most of survey data are regarded as "small", that does not fit into the "big data" category that we normally talk about in natural language processing or deep learning. For example, I just think that it is impossible to train a state-of-the-art neural NLP models like Transformers, RNN, LSTM from a pool of survey data.
However, I see several private companies which claim that they use deep learning methods to analyze survey data (for sentiment analysis, thematic analysis, etc). So I am just curious in what general manner deep learning models are applied in analysis of open-ended survey questions. (Certainly these private companies are not using anything like Transformer to do sentiment analysis on their client's survey data? or do these survey consulting companies tend to deal with very large survey data that is often not available in public?)
Thank you,
Have a look at https://static1.squarespace.com/static/51ed234ae4b0867e2385d879/t/54b327a7e4b034ff307ca427/1421027250851/structural-topic-models-for-open-ended-survey-responsesArticle Toward deep learning for adult students in online courses
.pdfArticle Content Analysis: Review of Methods and Their Applications i...
Had read an interesting article about open-ended survey question and the use of DL.
https://blog.perceptyx.com/open-ended-survey-questions-analysis
May be you could find something useful here.
- Badges
- Science topic
- Similar topics
- Statistics
- Survey