Using Transfer Learning To Classify Long Unstructured Texts with Small Amounts of Labeled Data.
Carlos Alberto Alvares RochaMarcos Vinícius Pinheiro DibLi WeigangAndrea Ferreira Portela NunesAllan Victor Almeida FariaDaniel Oliveira CajueiroMaísa Kely de MeloVictor Rafael Rezende CelestinoPublished in: WEBIST (2022)
Keyphrases
- labeled data
- learning to classify
- text classification
- transfer learning
- unlabeled data
- labeled data for training
- semi supervised learning
- semi supervised
- active learning
- domain adaptation
- text categorization
- supervised learning
- training data
- labeled training data
- labeled examples
- cross domain
- class labels
- semi supervised classification
- labeled and unlabeled data
- co training
- data points
- small number
- label propagation
- knowledge transfer
- learning tasks
- prior knowledge
- naive bayes
- learning algorithm
- feature selection
- text mining
- unsupervised learning
- labeling process
- training examples
- databases
- information retrieval
- labeled training set
- transferring knowledge
- neural network
- source domain
- natural language
- reinforcement learning
- text classifiers
- real world
- data sets