A Noisy-Channel Approach to Question Answering

By Abdessamad Echihabi, Daniel Marcu. 2003

171 citations recorded by the source corpus.

We introduce a probabilistic noisy-channel model for question answering and we show how it can be exploited in the context of an end-to-end QA system. Our noisy-channel system outperforms a state-of-the-art rule-based QA system that uses similar resources. We also show that the model we propose is flexible enough to accommodate within one mathematical framework many QA-specific resources and techniques, which range from the exploitation of WordNet, structured, and semi-structured databases to reasoning, and paraphrasing.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

Internet-augmented language models through few-shot prompting for open-domain question answering (2022, methodology): “…inference, where we choose an answer that maximizes p ( a | q ) = =1 p tfidf ( p i | q ) · p ( a i | q, p i ) , referred to as RAG throughout this work as it is inspired from the model of Lewis et al. [7] (ii) Noisy channel inference, where we choose an answer that maximizes p ( a i , q , [25, 26].”
Improving Machine Translation Performance by Exploiting Non-Parallel Corpora (2005, background): “Since then, they were found useful in many other NLP applications (e.g., word sense tagging [Diab and Resnik 2002] and question answering [Echihabi and Marcu 2003]).”
Learning to Rank Short Text Pairs with Convolutional Deep Neural Networks (2015, background): “2 can be viewed as a model of the noisy channel approach from machine translation, which has been widely used as a scoring model in information retrieval and question answering [13].”
What is the Jeopardy Model? A Quasi-Synchronous Grammar for QA (2007, background): “…developed powerful grammatical formalisms and statistical models for representing and learning these tree-to-tree relations (Wu and Wong, 1998; Eisner, 2003; Gildea, 2003; Melamed, 2004; Ding and Palmer, 2005; Quirk et al., 2005; Galley et al., 2006; Smith and Eisner, 2006, inter alia ).”
Alignment over Heterogeneous Embeddings for Question Answering (2019, methodology): “For instance, alignment approaches have received considerably less interest recently, despite their initial successes (Echihabi and Marcu, 2003; Surdeanu et al., 2011, inter alia).”
Data-Driven Response Generation in Social Media (2011, methodology): “Statistical machine translation has been applied to a sm¨org˚asbord of NLP problems, including question answering (Echihabi and Marcu, 2003), semantic parsing and generation (Wong and Mooney, 2006; Wong and Mooney, 2007), summarization (Daum´e III and Marcu, 2009), generating bid-phrases in on-line…”
Modeling Relational Information in Question-Answer Pairs with Convolutional Neural Networks (2016, methodology): “In this model, we seek a transformation of the candidate document x a = Mx a that is the closest to the input query x q .”
Semantic Text Matching for Long-Form Documents (2019, methodology): “In addition to the convolutional features, DeepQA uses a matching matrix [6] to measure an asymmetric similarity between the features of two documents with a noisy channel [14].”
Good Question! Statistical Ranking for Question Generation (2010, background): “…has focused on retrieval or extraction (e.g., Ravichan-dran and Hovy, 2001; Hovy et al., 2001), models of the transformation from answers to questions have also been developed (Echihabi and Marcu, 2003) with the goal of ﬁnding correct answers given a question (e.g., in a source-channel framework).”
Question Answering Using Enhanced Lexical Semantic Models (2013, background): “As any QA system can virtually be decomposed into two major high-level components, retrieval and selection (Echihabi and Marcu, 2003), the answer selection problem is clearly critical.”
Modelling input texts: from Tree Kernels to Deep Learning (2015, background): “4 can be viewed as a model of the noisy channel approach from machine translation, which has been widely used as a scoring model in information retrieval and question answering [42].”
Knowledge-Based Question Answering as Machine Translation (2014, background): “Echihabi and Marcu (2003) have developed a noisy-channel model for QA, which explains how a sentence containing an answer to a given question can be rewritten into that question through a sequence of stochastic operations.”