Improving Bi-encoder Document Ranking Models with Two Rankers and Multi-teacher Distillation.
Jaekeol ChoiEuna JungJangwon SuhWonjong RheePublished in: CoRR (2021)
Keyphrases
- learning to rank
- ranking models
- information retrieval
- learning to rank algorithms
- document retrieval
- retrieval systems
- loss function
- ranking functions
- ranking svm
- learning process
- keywords
- information retrieval systems
- normalized discounted cumulative gain
- benchmark datasets
- web search
- web documents
- data sets
- machine learning
- directly optimize
- feature extraction
- text documents
- document clustering
- supervised learning
- ranked list
- evaluation measures
- vector space model
- web mining
- clickthrough data
- learning algorithm
- language model
- document collections