The TREC-8 Question Answering Track Evaluation

By Ellen M. Voorhees, Dawn M. Tice. 2000

416 citations recorded by the source corpus.

How researchers see this paper

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

In Evaluation & Benchmarks

Citations in the corpus

Universal Language Model Fine-tuning for Text Classification (2018, methodology): “Question Classification We use the six-class version of the small TREC dataset (Voorhees and Tice, 1999) dataset of open-domain, fact-based questions divided into broad semantic categories.”
Learned in Translation: Contextualized Word Vectors (2017, methodology): “For question classification, we use the small TREC dataset [Voorhees and Tice, 1999] dataset of open-domain, fact-based questions divided into broad semantic categories.”
Fine-tuned Language Models for Text Classification (2018, methodology): “Classiﬁcation We use the six-class version of the small TREC dataset (Voorhees and Tice, 1999) dataset of open-domain, fact-based questions divided into broad semantic categories.”
Data Augmentation for BERT Fine-Tuning in Open-Domain Question Answering (2019, background): “This task definition, which combines a strong element of information retrieval, traces back to the Text Retrieval Conferences (TRECs) in the late 1990s (Voorhees and Tice, 1999), but there is a recent resurgence of interest in this formulation (Chen et al.”
WikiQA: A Challenge Dataset for Open-Domain Question Answering (2015, methodology): “The policy is adopted both by the official QASENT tracks (Voorhees and Tice, 1999) and by Wang et al.”
How to Fine-Tune BERT for Text Classification? (2019, neutral): “Question classiﬁcation For question classiﬁ-cation, we evaluate our method on the six-class version of the TREC dataset (Voorhees and Tice, 1999) and Yahoo!”
The TREC question answering track (2001, neutral): “The results of the TREC-8 assessment process demonstrated that assessors do have legitimate differences of opinion as to what constitutes an acceptable answer even for these deliberately constrained questions (Voorhees and Tice 2000b).”
A Replication Study of Dense Passage Retriever (2021, background): “Both the task formulation and the pipeline architecture for tackling the task date from the late 1990s (Voorhees and Tice, 1999), so this general approach has a long history that predates neural networks.”
Does an LSTM forget more than a CNN? An empirical study of catastrophic forgetting in NLP (2019, background): “• TREC Question classiﬁcation (“TREC”): coarse-grained classiﬁcation of questions, based on 6 classes (Voorhees and Tice, 1999).”
An Exploration of Approaches to Integrating Neural Reranking Models in Multi-Stage Ranking Architectures (2017, methodology): “Ultimately, the data derive from the Question Answering Tracks from TREC 8–13 [20, 21].”
A Machine Learning Approach to Answering Questions for Reading Comprehension Tests (2000, background): “One exception is the work of (Radev et al., 2000) at the TREC-8 QA track, which uses logistic regression to rank potential answers using a training set with seven features.”
Frontiers of biomedical text mining: current progress (2007, background): “them in context by providing supporting information and linking to original source documents [78].”
Distant Supervision for Multi-Stage Fine-Tuning in Retrieval-Based Question Answering (2020, neutral): “Our task definition, which combines a strong element of information retrieval, traces back to the Text Retrieval Conferences (TRECs) in the late 1990s [23], but there is a recent resurgence of interest in this formulation, starting with Chen et al.”
Answer Extraction from Passage Graph for Question Answering (2013, neutral): “• TREC [Voorhees, 1999] QA data set from the year 1999 to 2007.”
Building a question answering test collection (2000, background): “More details regarding question judging are given in the track report in the TREC-8 proceedings [16].”
Overview of the TREC 2003 Question Answering Track (2004, background): “Finally, extracts must refer to the famous entity itself and not imitations, copies, and the like to be responsive to questions about a famous entity (e.g., extracts about the Taj Mahal casino are not responsive to questions regarding “the Taj Mahal”) [ 6 ].”