Using Imperfect Surrogates for Downstream Inference: Design-based Supervised Learning for Social Science Applications of Large Language Models.
Naoki EgamiMusashi HinckBrandon M. StewartHanying WeiPublished in: NeurIPS (2023)
Keyphrases
- language model
- social sciences
- supervised learning
- language modeling
- probabilistic model
- statistical language models
- information retrieval
- document retrieval
- speech recognition
- language model for information retrieval
- digital archiving
- n gram
- artificial intelligence
- test collection
- context sensitive
- bayesian networks
- training data
- language modelling
- machine learning
- error rate
- training set
- digital government
- relevance model
- data mining