Neural Passage Quality Estimation for Static Pruning

• URL: http://arxiv.org/abs/2407.12170v1
• Date: Tue, 16 Jul 2024 20:47:54 GMT
• Title: Neural Passage Quality Estimation for Static Pruning
• Authors: Xuejun Chang, Debabrata Mishra, Craig Macdonald, Sean MacAvaney,
• Abstract summary: We explore whether neural networks can effectively predict which of a document's passages are unlikely to be relevant to any query submitted to the search engine. We find that our novel methods for estimating passage quality allow passage corpora to be pruned considerably. This work sets the stage for developing more advanced neural "learning-what-to-index" methods.
• Score: 23.662724916799004
• License:
• Abstract: Neural networks -- especially those that use large, pre-trained language models -- have improved search engines in various ways. Most prominently, they can estimate the relevance of a passage or document to a user's query. In this work, we depart from this direction by exploring whether neural networks can effectively predict which of a document's passages are unlikely to be relevant to any query submitted to the search engine. We refer to this query-agnostic estimation of passage relevance as a passage's quality. We find that our novel methods for estimating passage quality allow passage corpora to be pruned considerably while maintaining statistically equivalent effectiveness; our best methods can consistently prune >25% of passages in a corpora, across various retrieval pipelines. Such substantial pruning reduces the operating costs of neural search engines in terms of computing resources, power usage, and carbon footprint -- both when processing queries (thanks to a smaller index size) and when indexing (lightweight models can prune low-quality passages prior to the costly dense or learned sparse encoding step). This work sets the stage for developing more advanced neural "learning-what-to-index" methods.

Related papers

• Likelihood as a Performance Gauge for Retrieval-Augmented Generation [78.28197013467157]
  We show that likelihoods serve as an effective gauge for language model performance. We propose two methods that use question likelihood as a gauge for selecting and constructing prompts that lead to better performance.
  arXiv  Detail & Related papers  (2024-11-12T13:14:09Z)
• Fine-tuning -- a Transfer Learning approach [0.22344294014777952]
  Missingness in Electronic Health Records (EHRs) is often hampered by the abundance of missing data in this valuable resource. Existing deep imputation methods rely on end-to-end pipelines that incorporate both imputation and downstream analyses. This paper explores the development of a modular, deep learning-based imputation and classification pipeline.
  arXiv  Detail & Related papers  (2024-11-06T14:18:23Z)
• How Does Generative Retrieval Scale to Millions of Passages? [68.98628807288972]
  We conduct the first empirical study of generative retrieval techniques across various corpus scales. We scale generative retrieval to millions of passages with a corpus of 8.8M passages and evaluating model sizes up to 11B parameters. While generative retrieval is competitive with state-of-the-art dual encoders on small corpora, scaling to millions of passages remains an important and unsolved challenge.
  arXiv  Detail & Related papers  (2023-05-19T17:33:38Z)
• Curating corpora with classifiers: A case study of clean energy sentiment online [0.0]
  Large-scale corpora of social media posts contain broad public opinion. Surveys can be expensive to run and lag public opinion by days or weeks. We propose a method for rapidly selecting the best corpus of relevant documents for analysis.
  arXiv  Detail & Related papers  (2023-05-04T18:15:45Z)
• CorpusBrain: Pre-train a Generative Retrieval Model for Knowledge-Intensive Language Tasks [62.22920673080208]
  Single-step generative model can dramatically simplify the search process and be optimized in end-to-end manner. We name the pre-trained generative retrieval model as CorpusBrain as all information about the corpus is encoded in its parameters without the need of constructing additional index.
  arXiv  Detail & Related papers  (2022-08-16T10:22:49Z)
• A Transfer Learning Pipeline for Educational Resource Discovery with Application in Leading Paragraph Generation [71.92338855383238]
  We propose a pipeline that automates web resource discovery for novel domains. The pipeline achieves F1 scores of 0.94 and 0.82 when evaluated on two similar but novel target domains. This is the first study that considers various web resources for survey generation.
  arXiv  Detail & Related papers  (2022-01-07T03:35:40Z)
• Mining Implicit Relevance Feedback from User Behavior for Web Question Answering [92.45607094299181]
  We make the first study to explore the correlation between user behavior and passage relevance. Our approach significantly improves the accuracy of passage ranking without extra human labeled data. In practice, this work has proved effective to substantially reduce the human labeling cost for the QA service in a global commercial search engine.
  arXiv  Detail & Related papers  (2020-06-13T07:02:08Z)
• Learning the Travelling Salesperson Problem Requires Rethinking Generalization [9.176056742068813]
  End-to-end training of neural network solvers for graph optimization problems such as the Travelling Salesperson Problem (TSP) have seen a surge of interest recently. While state-of-the-art learning-driven approaches perform closely to classical solvers when trained on trivially small sizes, they are unable to generalize the learnt policy to larger instances at practical scales. This work presents an end-to-end neural optimization pipeline that unifies several recent papers in order to identify the principled biases, model architectures and learning algorithms that promote generalization to instances larger than those seen in training.
  arXiv  Detail & Related papers  (2020-06-12T10:14:15Z)
• WoodFisher: Efficient Second-Order Approximation for Neural Network Compression [35.45199662813043]
  We develop a method to compute a faithful and efficient estimate of the inverse Hessian. Our main application is to neural network compression. We show how our method can be extended to take into account first-order information.
  arXiv  Detail & Related papers  (2020-04-29T17:14:23Z)
• Binary Neural Networks: A Survey [126.67799882857656]
  The binary neural network serves as a promising technique for deploying deep models on resource-limited devices. The binarization inevitably causes severe information loss, and even worse, its discontinuity brings difficulty to the optimization of the deep network. We present a survey of these algorithms, mainly categorized into the native solutions directly conducting binarization, and the optimized ones using techniques like minimizing the quantization error, improving the network loss function, and reducing the gradient error.
  arXiv  Detail & Related papers  (2020-03-31T16:47:20Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.