Pre-trained Language Model based Ranking in Baidu Search

Abstract

As the heart of a search engine, the ranking system plays a crucial role in satisfying users' information demands. More recently, neural rankers fine-tuned from pre-trained language models (PLMs) establish state-of-the-art ranking effectiveness. However, it is nontrivial to directly apply these PLM-based rankers to the large-scale web search system due to the following challenging issues:(1) the prohibitively expensive computations of massive neural PLMs, especially for long texts in the web-document, prohibit their deployments in an online ranking system that demands extremely low latency;(2) the discrepancy between existing ranking-agnostic pre-training objectives and the ad-hoc retrieval scenarios that demand comprehensive relevance modeling is another main barrier for improving the online ranking system;(3) a real-world search engine typically involves a committee of ranking components, and thus the compatibility of the individually fine-tuned ranking model is critical for a cooperative ranking system. In this work, we contribute a series of successfully applied techniques in tackling these exposed issues when deploying the state-of-the-art Chinese pre-trained language model, i.e., ERNIE, in the online search engine system. We first articulate a novel practice to cost-efficiently summarize the web document and contextualize the resultant summary content with the query using a cheap yet powerful Pyramid-ERNIE architecture. Then we endow an innovative paradigm to finely exploit the large-scale noisy and biased post-click behavioral data for relevance-oriented pre-training. We also propose a human-anchored fine-tuning strategy tailored for the online ranking system, aiming to stabilize the ranking signals across various online components. Extensive offline and online experimental results show that the proposed techniques significantly boost the search engine's performance.

9-pages, 3 figures, 7 tables, SIGKDD 2021 accepted paper

References in corpus (23)

• Distilling the Knowledge in a Neural Network
• BioBERT: a pre-trained biomedical language representation model for biomedical text mining
• Longformer: The Long-Document Transformer
• ALBERT: A Lite BERT for Self-supervised Learning of Language Representations
• Linformer: Self-Attention with Linear Complexity
• A Deep Relevance Matching Model for Ad-hoc Retrieval
• ERNIE: Enhanced Representation through Knowledge Integration
• Quantizing deep convolutional networks for efficient inference: A whitepaper
• End-to-End Neural Ad-hoc Ranking with Kernel Pooling
• Fine-Tuning Language Models from Human Preferences
• Passage Re-ranking with BERT
• Multi-Stage Document Ranking with BERT
• ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT
• Neural Interactive Collaborative Filtering
• Rethinking Attention with Performers
• Pretrained Transformers for Text Ranking: BERT and Beyond
• Reinforcement Learning to Optimize Long-term User Engagement in Recommender Systems
• Better Fine-Tuning by Reducing Representational Collapse
• Modeling of Pruning Techniques for Deep Neural Networks Simplification
• Deep Relevance Ranking Using Enhanced Document-Query Interactions
• Multi-Stage Pre-training for Low-Resource Domain Adaptation
• Pre-training Text-to-Text Transformers for Concept-centric Common Sense
• Pre-trained Language Model for Web-scale Retrieval in Baidu Search