Generating embeddings through doing L2-normalization to sgns embeddings trained by all data (train,dev,test).
- SQLNet: Generating Structured Queries from Natural Language without Reinforcement Learning
- Coarse-to-Fine Decoding for Neural Semantic Parsing*
- Simplification Using Paraphrases and Context-based Lexical Substitution
- Entity Matching on Web Tables: a Table Embeddings Approach for Blocking
- Table2Vec: Neural Word and Entity Embeddings for Table Population and Retrieval
An Ensemble Method to Produce High-Quality Word Embeddings
combine conceptNetwith word2vec and glove; SVDImproving Distributional Similarity with Lessons Learned from Word Embeddings
combine predictive embedding models with count based modelsColNet: Embedding the Semantics of Web Tables for Column Type Prediction
CNN and transfer learning to deal with words did not appear at KBSentence Modeling via Multiple Word Embeddings and Multi-level Comparison for Semantic Textual Similarity
Structural Embedding of Syntactic Trees for Machine Comprehension
Supervised Learning of Universal Sentence Representations from Natural Language Inference Data
Retrofitting Word Vectors to Semantic Lexicons
ConceptNet 5.5: An Open Multilingual Graph of General Knowledge
Deep contextualized word representations(ELMo)
Training Relation Embeddings under Logical Constraints
A La Carte Embedding: Cheap but Effective Induction of Semantic Feature Vectors
Discussion on different embeddings
Next to do: fix k8s container not found bug