2020
DOI: 10.1017/s1351324920000017
|View full text |Cite
|
Sign up to set email alerts
|

Constrained BERT BiLSTM CRF for understanding multi-sentence entity-seeking questions

Abstract: We present the novel task of understanding multi-sentence entity-seeking questions (MSEQs), that is, the questions that may be expressed in multiple sentences, and that expect one or more entities as an answer. We formulate the problem of understanding MSEQs as a semantic labeling task over an open representation that makes minimal assumptions about schema or ontology-specific semantic vocabulary. At the core of our model, we use a BiLSTM (bidirectional LSTM) conditional random field (CRF), and to overcome the… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
9
0

Year Published

2020
2020
2022
2022

Publication Types

Select...
3
3
1

Relationship

1
6

Authors

Journals

citations
Cited by 13 publications
(9 citation statements)
references
References 48 publications
0
9
0
Order By: Relevance
“…(1) Frequency-based methods [1,2]; (2) Rule/template-based methods [3,4]; (3) Graph theory-based methods [6][7][8]; (4) Based on CRF or combined CRF and deep learning methods [5,[9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25].…”
Section: Overview Of the Methods Of Opinion Targets Extractionmentioning
confidence: 99%
See 1 more Smart Citation
“…(1) Frequency-based methods [1,2]; (2) Rule/template-based methods [3,4]; (3) Graph theory-based methods [6][7][8]; (4) Based on CRF or combined CRF and deep learning methods [5,[9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25].…”
Section: Overview Of the Methods Of Opinion Targets Extractionmentioning
confidence: 99%
“…Hu [14] designed a bidirectional long short-term memory (BiLSTM)-CRF model to achieve opinion extraction of product reviews. Danish [15] proposed a bidirectional encoder representations from transformers (BERT)-BiLSTM-CRF model for entity-seeking of multi-sentence. Huang [16] also used the BiLSTM-CRF model for sequence labeling tasks.…”
Section: Overview Of the Methods Of Opinion Targets Extractionmentioning
confidence: 99%
“…Our work builds on the recently-released POI entity-recommendation QA task (Contractor et al, 2019(Contractor et al, , 2020. Two approaches have been developed for this task: semantic parsing of unstructured user questions to query a semi-structured knowledge store (Contractor et al, 2020), and an end-to-end trainable neural model operating over a corpus of unstructured reviews to represent POIs. Neither of these approaches explicitly reason on spatial constraints, even though the questions contain them.…”
Section: Related Workmentioning
confidence: 99%
“…In order to get mentions of locations in questions, we manually label a set of 425 questions from the training set for location mentions. We then use a BERT-BiLSTM CRF (Contractor et al, 2020) based tagger trained on this set to label locations. We report some experiments using this model referred to as CSQA and compare it with CSRQA and spatio-textual CSRQA.…”
Section: A2 Location Taggermentioning
confidence: 99%
“…Answers to these questions are usually names of entities such as the name of a restaurant. We are building methods that utilize large text collections of entity reviews to answer such questions automatically [5]. The work involves being able to understand a user's preference or constraints as well as distill knowledge from the vast collection of reviews describing each of those entities [6].…”
Section: What Project Are You Currently Leading That Have the Potentimentioning
confidence: 99%