Comput. Linguistics 2022

17 papers
In 2022, the journal Computational Linguistics published 20 papers focusing on advancements in neural architectures and multilingual processing. Key research directions included the development of robust neural machine translation (NMT) systems, particularly for long sequences, and the application of Monte Carlo dropout for uncertainty quantification. The year's contributions also addressed the specific challenges of NMT for short texts and explored hierarchical interpretations of neural text classification models. Notable work included surveys on deep learning for text style transfer and investigations into the boundary between automatic speech recognition (ASR) and natural language understanding (NLU) within conversational dialog systems. Research also delved into dependency parsing through nucleus composition and typology-based language adapters for multilingual sequence labeling. Prolific contributors to the volume included Baosong Yang, Haibo Zhang, Liang Yao, and Iryna Gurevych. Other significant topics covered neural query language identification and the implementation of annotation curricula to train non-expert annotators.

Top Research Directions

The 17 research directions with Comput. Linguistics 2022 papers, ranked by paper count, next to 2021. Share is the fraction of the edition’s papers in the direction.

#Direction20222021Share 2022Share 20212021 rank
1Back-Translation for Synthetic Data Augmentation105.9%0.0%—
2Bidirectional Co-attention for Joint Sequence Representation105.9%0.0%—
3Causal Interventions for Neural Representation Analysis105.9%0.0%—
4Disentangling Content and Style in Latent Representations105.9%0.0%—
5Disentangling Representation Knowledge from Probe Learning105.9%0.0%—
6Estimating Model Uncertainty via Stochastic Forward Passes105.9%0.0%—
7Linear Probing of Internal Network Representations105.9%0.0%—
8Modeling Language through Structured Semantic Frames105.9%0.0%—
9Resolving Query Underspecification through Interactive Clarification105.9%0.0%—
10Segment Isolated Attention Masking105.9%0.0%—
11Style Transfer without Parallel Corpora105.9%0.0%—
12The Complexity Gap in Artificial Language Modeling105.9%0.0%—
13The Compositional Generalization Gap in Neural Networks105.9%0.0%—
14The Evaluation Gap in Text Style Transfer105.9%0.0%—
15Training Models via Progressive Task Complexity105.9%0.0%—
16Uncertainty Calibration for Large Language Models105.9%0.0%—
17Why Neural Networks Fail to Generalize Structural Recursion105.9%0.0%—

Download all directions as CSV

Institutions

Institutions credited by Comput. Linguistics 2022 papers: a paper credits every institution named in its affiliations, once. 23 institutions share 13 credited papers.

#InstitutionPapersShare20212021 rank
1Technische Universität Darmstadt 215.4%116
203dgzv892 17.7%0—
3Alibaba Group (China) 17.7%0—
4Amazon (United States) 17.7%0—
5ETH Zurich 17.7%0—
6Google (United States) 17.7%21
7Ludwig-Maximilians-Universität München 17.7%0—
8Max Planck Society 17.7%0—
9Norwegian Computing Center 17.7%0—
10Technion – Israel Institute of Technology 17.7%111
11The Alan Turing Institute 17.7%0—
12UNESCO 17.7%0—
13Universitat Rovira i Virgili 17.7%0—
14University of California San Diego 17.7%0—
15University of Macau 17.7%0—
16University of Michigan 17.7%0—
17University of Oslo 17.7%0—
18University of Toronto 17.7%0—
19University of Warwick 17.7%0—
20University of Waterloo 17.7%0—
21Utrecht University 17.7%0—
22Vector Institute 17.7%0—
23York University 17.7%115

Download all institutions as CSV

About this data

Papers. Comput. Linguistics 2022 lists 17 papers on Lacuna, duplicate versions counted once; rejected and withdrawn submissions are left out.

Directions. Directions are Lacuna’s clustering of the concept elements extracted from each paper, so a paper can belong to several and direction counts add up to more than the paper count. A direction’s share is its papers divided by the edition’s papers. Comparisons with 2021 use shares, since the edition went from 17 to 17 papers.

Institutions. A paper credits each institution named in its affiliations, resolved to a ROR record, once. 13 of 17 papers (76.5%) credit at least one institution and are the base for institution shares; the rest have no resolved affiliation yet.

Data. Counts come from Lacuna’s data build of 2026-11-20 and change as papers are added. Tables as CSV: directions · institutions. Cite as: Lacuna, “Comput. Linguistics 2022”, https://lacuna.tiptreesystems.com/venue/comput-linguistics/2022.

Top Authors

AuthorPapers
Baosong Yang2
Iryna Gurevych2
Ji-Ung Lee2
Ahmet Üstün1
Alexandra Birch1
Antonio Valerio Miceli Barone1
Barry Haddow1
Boxing Chen1
Dayiheng Liu1
Denis Paperno1
Derek F. Wong1
Di Jin1
Dilek Hakkani Tur1
Enamul Hoque1
Hanqi Yan1
Hinrich Schuetze1
Jimmy Huang1
Joakim Nivre1
Lidia S. Chao1
Lilja Øvrelid1

Impactful Papers

Papers from Comput. Linguistics 2022 that shaped research clusters.

Foundational

Papers

Title
Annotation Curricula to Implicitly Train Non-Expert Annotators
Challenges of Neural Machine Translation for Short Texts
Deep Learning for Text Style Transfer: A Survey
Domain Adaptation with Pre-trained Transformers for Query-Focused Abstractive Text Summarization
Effective Approaches to Neural Query Language Identification
Erratum: Annotation Curricula to Implicitly Train Non-Expert Annotators
Hierarchical Interpretation of Neural Text Classification
Noun2Verb: Probabilistic Frame Semantics for Word Class Conversion
Nucleus Composition in Transition-based Dependency Parsing
On Learning Interpreted Languages with Recurrent Models
Position Information in Transformers: An Overview
Probing Classifiers: Promises, Shortcomings, and Advances
Revise and Resubmit: An Intertextual Model of Text-based Collaboration in Peer Review
Revisiting the Boundary between ASR and NLU in the Age of Conversational Dialog Systems
Survey of Low-Resource Machine Translation
The Text Anonymization Benchmark (TAB): A Dedicated Corpus and Evaluation Framework for Text Anonymization
UDapter: Typology-based Language Adapters for Multilingual Dependency Parsing and Sequence Labeling