The overall historical behaviors are various but noisy while search behaviors are always sparse.
Recently, the development of pre-trained language models has brought natural language processing (NLP) tasks to the new state-of-the-art.
We do the experiment on our own text classification dataset, which is manually labeled, because we re-label the noisy data in our dataset for our industry application.
We present a simple methods to leverage the table content for the BERT-based model to solve the text-to-SQL problem.
Ranked #1 on Code Generation on WikiSQL
We do the experiments on the semantic textual similarity dataset, Quora Question Pairs, and process the dataset for sentence ranking.