---
title: "Rochester — research institutions and their papers"
description: "Rochester (United States): 2 institutions with papers on Elamur Research — 2 universities — and 11,106 citations of their works."
canonical: "https://research.elamur.ai/free/en/agglomerations/rochester-us/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/agglomerations/rochester-us/)

[home](https://md.research.elamur.ai/free/en/) / [institutions](https://md.research.elamur.ai/free/en/orgs/) / [agglomerations](https://md.research.elamur.ai/free/en/agglomerations/) / Rochester

# Rochester

Agglomeration [United States](https://md.research.elamur.ai/free/en/countries/united-states/) · metro population 848,028 (functional urban area, GHSL 2015)

**2** institutions with a page**29** works in our library**11,106** citations of their works

## About

### Profile

Rochester is a city region in the [United States](https://md.research.elamur.ai/free/en/countries/united-states/). On this site its research comes from two universities: the [University of Rochester](https://md.research.elamur.ai/free/en/universities/university-of-rochester/) and the [Rochester Institute of Technology](https://md.research.elamur.ai/free/en/universities/rochester-institute-of-technology/). The University of Rochester has the larger body of work here and the more cited one. The Rochester Institute of Technology has fewer works on the site, but several of them are widely cited.

Both institutions publish mainly in natural language processing and nearby parts of artificial intelligence. Their work here covers several areas:

semantic annotation, and meaning representations for generation and translation; - vision and language, and multimodal learning; - natural language inference; - language identification, offensive language detection and lexical complexity; - resources for low-resource languages; - continual learning and the evaluation of large models.

No companies or research institutes from the region appear on this site. The region's science here is therefore the science of these two universities.

### History

The oldest work here tied to the region is [The Proposition Bank](https://md.research.elamur.ai/free/en/papers/palmer-2005-proposition-bank-annotated-corpus-semantic/) from 2005, associated with the University of Rochester. It adds predicate-argument labels to the Wall Street Journal portion of the Penn Treebank. It is also the most cited work connected with Rochester on this site. Its focus on semantic roles and annotated corpora sets a theme that runs through much of the later work: describing what sentences mean in a way that machines can learn from.

The next works come in 2016, and they mark a turn toward vision and language at the University of Rochester. [Visual Storytelling](https://md.research.elamur.ai/free/en/papers/huang-2016-visual-storytelling/) builds a dataset for producing sequences of language from sequences of images. [Generating Natural Questions About an Image](https://md.research.elamur.ai/free/en/papers/mostafazadeh-2016-generating-natural-questions-about-image/) defines the task of visual question generation.

Between 2018 and 2019 the University of Rochester's work on meaning representations becomes visible. [A Graph-to-Sequence Model for AMR-to-Text Generation](https://md.research.elamur.ai/free/en/papers/song-2018-graph-sequence-model-amr-text/) appears in 2018, and [Semantic Neural Machine Translation Using AMR](https://md.research.elamur.ai/free/en/papers/song-2019-semantic-neural-machine-translation-amr/) follows in 2019. Natural language inference resources also appear in 2018. In the same years the Rochester Institute of Technology appears on the site with a 2018 review of continual learning and a 2019 survey of language identification.

From 2020 onward the region's output broadens. Work on multimodal transformers arrives in 2020. There are also shared-task reports and new datasets for languages other than English. Bangla and African language resources follow in 2022. The most recent work here is a 2025 benchmark for frontier language models.

### Research profile

Semantic annotation and meaning representation form the oldest line among the works here. The Proposition Bank supplies numbered argument labels for verbs, and its reported labelling experiments reach high accuracy.

The AMR work at the University of Rochester builds on graph-based meaning. The 2018 generation paper swaps a linear encoder for a graph-state LSTM, so that text is produced directly from the structure of an AMR graph. The 2019 translation paper then tests whether giving a neural translation system the meaning graph of the input sentence helps, and it reports gains in BLEU.

Vision and language is a second strand at the University of Rochester. The visual storytelling dataset, called SIND, is presented as the first resource for sequential vision-to-language. The visual question generation paper asks systems to pose natural, engaging questions about images rather than to describe them.

Multimodal learning continues this direction in [Integrating Multimodal Information in Large Pretrained Transformers](https://md.research.elamur.ai/free/en/papers/rahman-2020-integrating-multimodal-information-large-pretrained/). That paper proposes the Multimodal Adaptation Gate. It is a light attachment that lets a pretrained language model take in nonverbal signals without changing its original structure.

Inference and sentence representation are the focus of [Collecting Diverse Natural Language Inference Problems for Sentence Representation](https://md.research.elamur.ai/free/en/papers/poliak-2018-collecting-diverse-natural-language-inference/). It recasts existing resources into inference problems, so that particular reasoning abilities can be tested cheaply and in a form that can be extended.

Low-resource language work at the University of Rochester includes [BanglaBERT](https://md.research.elamur.ai/free/en/papers/bhattacharjee-2022-banglabert-language-model-pretraining-benchmarks/). It addresses Bangla, a language with a very large speaker base but few NLP resources, by providing both a pretrained model and benchmarks.

At the Rochester Institute of Technology, the works here focus on multilingual and applied language processing:

[Automatic Language Identification in Texts: A Survey](https://md.research.elamur.ai/free/en/papers/jauhiainen-2019-automatic-language-identification-texts/) reviews methods for deciding which language a text, or part of a text, is written in. It draws on a very long reference list. - [Offensive Language Identification in Greek](https://md.research.elamur.ai/free/en/papers/pitenis-2020-offensive-language-identification-greek/) creates an annotated dataset for abusive content in Greek, which previously had none. - [SemEval-2021 Task 1: Lexical Complexity Prediction](https://md.research.elamur.ai/free/en/papers/shardlow-2021-semeval-2021-task-1-lexical/) replaces the binary framing of complex word identification with continuous prediction of how hard a word is. - [MasakhaNER 2.0](https://md.research.elamur.ai/free/en/papers/adelani-2022-masakhaner-2-0-africa-centric/) provides a human-annotated news corpus for named entity recognition in many Sub-Saharan African languages, and studies transfer learning across them.

Machine translation evaluation appears through [Findings of the 2020 Conference on Machine Translation (WMT20)](https://md.research.elamur.ai/free/en/papers/barrault-2020-findings-2020-conference-machine-translation/). This report covers shared tasks held online alongside EMNLP 2020.

Learning and evaluation more broadly are represented by [Continual Lifelong Learning with Neural Networks: A Review](https://md.research.elamur.ai/free/en/papers/parisi-2018-continual-lifelong-learning-neural-networks/). It examines how networks trained on a stream of data tend to forget what they learned earlier.

The newest line is [Humanity's Last Exam](https://md.research.elamur.ai/free/en/papers/phan-2025-humanity-s-last-exam/). It responds to the near-saturation of benchmarks such as MMLU with a harder test of model capability.

### Leading scientists

At the University of Rochester the first person named is [Daniel Gildea](https://md.research.elamur.ai/free/en/authors/daniel-gildea/), whose work here belongs to the university's language-processing research. Next is [James F. Allen](https://md.research.elamur.ai/free/en/authors/james-f-allen/), also part of the university's natural language research.

[Linfeng Song](https://md.research.elamur.ai/free/en/authors/linfeng-song/) is the author named on the AMR-to-text generation paper and on the AMR-based translation paper. Together these form the graph-based meaning strand here. [Nasrin Mostafazadeh](https://md.research.elamur.ai/free/en/authors/nasrin-mostafazadeh/) is the first author of the visual question generation paper, which belongs to the vision-and-language line.

Further names at the university are [Joel Tetreault](https://md.research.elamur.ai/free/en/authors/joel-tetreault/), [Ehsan Hoque](https://md.research.elamur.ai/free/en/authors/ehsan-hoque/), [Aaron Steven White](https://md.research.elamur.ai/free/en/authors/aaron-steven-white/) and [Tagyoung Chung](https://md.research.elamur.ai/free/en/authors/tagyoung-chung/). Their work here falls within the university's research on language, multimodal signals and meaning.

At the Rochester Institute of Technology the first person named is [Marcos Zampieri](https://md.research.elamur.ai/free/en/authors/marcos-zampieri/). The institute's works here on language identification, offensive language and lexical complexity fall in the multilingual area this group represents. [Hakyung Sung](https://md.research.elamur.ai/free/en/authors/hakyung-sung/) follows.

The institute's people also include [Yi Wang](https://md.research.elamur.ai/free/en/authors/yi-wang/), [Qifan Wang](https://md.research.elamur.ai/free/en/authors/qifan-wang/), [Aman Chadha](https://md.research.elamur.ai/free/en/authors/aman-chadha/), [Salman Khan](https://md.research.elamur.ai/free/en/authors/salman-khan/), [Yidong Chen](https://md.research.elamur.ai/free/en/authors/yidong-chen/) and [Emily Prud'hommeaux](https://md.research.elamur.ai/free/en/authors/emily-prud-hommeaux/). This range of names matches a body of work that extends from shared tasks in language processing to large-model evaluation and continual learning.

### Landmark works

The most cited work linked to Rochester on this site is The Proposition Bank. Its semantic role labels on top of a widely used treebank made it a standard training and evaluation resource for semantic role labelling.

The WMT20 findings report and Humanity's Last Exam, both connected with the Rochester Institute of Technology, follow it in citations among the works here. The first summarises a yearly evaluation of translation systems. The second offers a new, harder benchmark at a time when older tests no longer separate strong models well.

Close behind are the continual learning review and the visual storytelling dataset. The review organises approaches to catastrophic forgetting. The dataset opened the problem of telling a story from a sequence of photos.

The graph-to-sequence AMR paper and BanglaBERT also stand out among the University of Rochester's works. Among the Rochester Institute of Technology's works, the language identification survey, MasakhaNER 2.0, the lexical complexity task and the Greek offensive language dataset each draw wide citation. The inference collection, the multimodal adaptation gate, the visual question generation paper and the AMR translation study complete the core of the region's most cited research here.

### Labs and units

The region's research here belongs to its two universities. The University of Rochester covers semantic resources, graph-based generation and translation, vision and language, multimodal transformers, inference and Bangla language modelling. The Rochester Institute of Technology covers multilingual resources and shared tasks, surveys of language identification and continual learning, and benchmark construction. No finer departments or laboratories are given for either university on this site.

### Collaborations

Many of the region's works are collaborative by design. The WMT20 findings and the SemEval-2021 lexical complexity task are reports from shared tasks that bring many teams together. MasakhaNER 2.0 comes from a community effort to build resources across African languages. Humanity's Last Exam gathers questions from a broad body of contributors. On the University of Rochester side, the Bangla work and the recasting of inference datasets also draw on resources and partners from beyond the university.

The two universities share common ground in natural language processing. One leans toward semantics and translation, the other toward multilingual resources and evaluation. Their works here do not show a direct joint paper between them. Links between industry and academia in the region cannot be traced here, since no Rochester companies appear on this site.

### Timeline

In 2005 the region's work here begins with the Proposition Bank at the University of Rochester. It makes semantic role annotation the earliest line.

In 2016 vision and language arrives at the University of Rochester, with the visual storytelling dataset and visual question generation.

In 2018 and 2019 the University of Rochester publishes its graph-to-sequence AMR generation model, its AMR-based translation study and its diverse inference collection. The Rochester Institute of Technology appears with its continual learning review and its language identification survey.

In 2020 and 2021 the multimodal adaptation gate appears at the University of Rochester. The Rochester Institute of Technology contributes the WMT20 findings, the Greek offensive language dataset and the SemEval-2021 lexical complexity task.

In 2022 both universities turn to languages with few resources: BanglaBERT at the University of Rochester and MasakhaNER 2.0 at the Rochester Institute of Technology.

In 2025 Humanity's Last Exam, connected with the Rochester Institute of Technology, carries the region's work into the evaluation of frontier language models.

## Institutions (2)

2 universities.

### 🎓 Universities (2)

1. **[University of Rochester](https://md.research.elamur.ai/free/en/universities/university-of-rochester/)**University · Rochester, [United States](https://md.research.elamur.ai/free/en/countries/united-states/)*22 works · 6,510 citations*
2. **[Rochester Institute of Technology](https://md.research.elamur.ai/free/en/universities/rochester-institute-of-technology/)**University · Rochester, [United States](https://md.research.elamur.ai/free/en/countries/united-states/)*7 works · 4,596 citations*

## Works in our library (29)

Works of its institutions, the most cited first.

1. [**Continual Lifelong Learning with Neural Networks: A Review**German I. Parisi, Ronald Kemker, Jose L. Part, et al.*2018 · 3,210 citations*](https://md.research.elamur.ai/free/en/papers/parisi-2018-continual-lifelong-learning-neural-networks/)
2. [**The Proposition Bank: An Annotated Corpus of Semantic Roles**Martha Palmer, Daniel Gildea, Paul Kingsbury*2005 · 3,208 citations*](https://md.research.elamur.ai/free/en/papers/palmer-2005-proposition-bank-annotated-corpus-semantic/)
3. [**Integrating Multimodal Information in Large Pretrained Transformers**Wasifur Rahman, Md Kamrul Hasan, Sangwu Lee, et al.*2020 · 742 citations*](https://md.research.elamur.ai/free/en/papers/rahman-2020-integrating-multimodal-information-large-pretrained/)
4. [**Findings of the 2020 Conference on Machine Translation (WMT20)**Loïc Barrault, Magdalena Biesialska, Ondřej Bojar, et al.*2020 · 401 citations*](https://md.research.elamur.ai/free/en/papers/barrault-2020-findings-2020-conference-machine-translation/)
5. [**Visual Storytelling**Ting-Hao Kenneth Huang, Francis Ferraro, Nasrin Mostafazadeh, et al.*2016 · 344 citations*](https://md.research.elamur.ai/free/en/papers/huang-2016-visual-storytelling/)
6. [**Generating Natural Questions About an Image**Nasrin Mostafazadeh, Ishan Misra, Jacob Devlin, et al.*2016 · 339 citations*](https://md.research.elamur.ai/free/en/papers/mostafazadeh-2016-generating-natural-questions-about-image/)
7. [**A Graph-to-Sequence Model for AMR-to-Text Generation**Linfeng Song, Yue Zhang, Zhiguo Wang, et al.*2018 · 329 citations*](https://md.research.elamur.ai/free/en/papers/song-2018-graph-sequence-model-amr-text/)
8. [**BanglaBERT: Language Model Pretraining and Benchmarks for Low-Resource Language Understanding Evaluation in Bangla**Abhik Bhattacharjee, Tahmid Hasan, Wasi Ahmad, et al.*2022 · 296 citations*](https://md.research.elamur.ai/free/en/papers/bhattacharjee-2022-banglabert-language-model-pretraining-benchmarks/)
9. [**Automatic Language Identification in Texts: A Survey**Tommi Jauhiainen, Marco Lui, Marcos Zampieri, et al.*2019 · 277 citations*](https://md.research.elamur.ai/free/en/papers/jauhiainen-2019-automatic-language-identification-texts/)
10. [**Humanity's Last Exam**Long Phan, Alice Gatti, Ziwen Han, et al.*2025 · 235 citations*](https://md.research.elamur.ai/free/en/papers/phan-2025-humanity-s-last-exam/)
11. [**Collecting Diverse Natural Language Inference Problems for Sentence Representation Evaluation**Adam Poliak, Aparajita Haldar, Rachel Rudinger, et al.*2018 · 190 citations*](https://md.research.elamur.ai/free/en/papers/poliak-2018-collecting-diverse-natural-language-inference/)
12. [**Offensive Language Identification in Greek**Zesis Pitenis, Marcos Zampieri, Tharindu Ranasinghe*2020 · 168 citations*](https://md.research.elamur.ai/free/en/papers/pitenis-2020-offensive-language-identification-greek/)
13. [**Semantic Neural Machine Translation Using AMR**Linfeng Song, Daniel Gildea, Yue Zhang, et al.*2019 · 159 citations*](https://md.research.elamur.ai/free/en/papers/song-2019-semantic-neural-machine-translation-amr/)
14. [**SemEval-2021 Task 1: Lexical Complexity Prediction**Matthew Shardlow, Richard Evans, Gustavo Henrique Paetzold, et al.*2021 · 155 citations*](https://md.research.elamur.ai/free/en/papers/shardlow-2021-semeval-2021-task-1-lexical/)
15. [**MasakhaNER 2.0: Africa-centric Transfer Learning for Named Entity Recognition**David Ifeoluwa Adelani, Graham Neubig, Sebastian Ruder, et al.*2022 · 150 citations*](https://md.research.elamur.ai/free/en/papers/adelani-2022-masakhaner-2-0-africa-centric/)
16. [**GLUCOSE: GeneraLized and COntextualized Story Explanations**Nasrin Mostafazadeh, Aditya Kalyanpur, Lori Moon, et al.*2020 · 144 citations*](https://md.research.elamur.ai/free/en/papers/mostafazadeh-2020-glucose-generalized-contextualized-story-explanations/)
17. [**A Corpus-Based Evaluation of Centering and Pronoun Resolution**Joel R. Tetreault*2001 · 125 citations*](https://md.research.elamur.ai/free/en/papers/tetreault-2001-corpus-based-evaluation-centering-pronoun/)
18. [**Neural Models of Factuality**Rachel Rudinger, Aaron Steven White, Benjamin Van Durme*2018 · 111 citations*](https://md.research.elamur.ai/free/en/papers/rudinger-2018-neural-models-factuality/)
19. [**AMR-to-text Generation with Synchronous Node Replacement Grammar**Linfeng Song, Xiaochang Peng, Yue Zhang, et al.*2017 · 108 citations*](https://md.research.elamur.ai/free/en/papers/song-2017-amr-text-generation-synchronous-node/)
20. [**Binarization of Synchronous Context-Free Grammars**Liang Huang, Hao Zhang, Daniel Gildea, et al.*2009 · 84 citations*](https://md.research.elamur.ai/free/en/papers/huang-2009-binarization-synchronous-context-free-grammars/)
21. [**Extracting and evaluating general world knowledge from the Brown Corpus**Lenhart Schubert, Matthew Tong*2003 · 75 citations*](https://md.research.elamur.ai/free/en/papers/schubert-2003-extracting-evaluating-general-world-knowledge/)
22. [**Two Constraints on Speech Act Ambiguity**Elizabeth A. Hinkelman, James F. Allen*1989 · 60 citations*](https://md.research.elamur.ai/free/en/papers/hinkelman-1989-two-constraints-speech-act-ambiguity/)
23. [**Analysis of Syntax-Based Pronoun Resolution Methods**Joel R. Tetreault*1999 · 60 citations*](https://md.research.elamur.ai/free/en/papers/tetreault-1999-analysis-syntax-based-pronoun-resolution/)
24. [**Temporal Reasoning in Natural Language Inference**Siddharth Vashishtha, Adam Poliak, Yash Kumar Lal, et al.*2020 · 57 citations*](https://md.research.elamur.ai/free/en/papers/vashishtha-2020-temporal-reasoning-natural-language-inference/)
25. [**Rhetorical Structure in Dialog**Amanda Stent*2000 · 56 citations*](https://md.research.elamur.ai/free/en/papers/stent-2000-rhetorical-structure-dialog/)

[See all 29 works →](https://md.research.elamur.ai/free/en/agglomerations/rochester-us/works/)

## Works by year

### 2020s

[**2026** 1 work](https://md.research.elamur.ai/free/en/papers/azad-2026-harf-speech-clinically-aligned-framework/) · [**2025** 1 work](https://md.research.elamur.ai/free/en/papers/phan-2025-humanity-s-last-exam/) · [**2022** 2 works](https://md.research.elamur.ai/free/en/agglomerations/rochester-us/year/2022/) · [**2021** 2 works](https://md.research.elamur.ai/free/en/agglomerations/rochester-us/year/2021/) · [**2020** 6 works](https://md.research.elamur.ai/free/en/agglomerations/rochester-us/year/2020/)

### 2010s

[**2019** 2 works](https://md.research.elamur.ai/free/en/agglomerations/rochester-us/year/2019/) · [**2018** 4 works](https://md.research.elamur.ai/free/en/agglomerations/rochester-us/year/2018/) · [**2017** 1 work](https://md.research.elamur.ai/free/en/papers/song-2017-amr-text-generation-synchronous-node/) · [**2016** 2 works](https://md.research.elamur.ai/free/en/agglomerations/rochester-us/year/2016/)

### 2000s

[**2009** 1 work](https://md.research.elamur.ai/free/en/papers/huang-2009-binarization-synchronous-context-free-grammars/) · [**2005** 1 work](https://md.research.elamur.ai/free/en/papers/palmer-2005-proposition-bank-annotated-corpus-semantic/) · [**2003** 1 work](https://md.research.elamur.ai/free/en/papers/schubert-2003-extracting-evaluating-general-world-knowledge/) · [**2001** 1 work](https://md.research.elamur.ai/free/en/papers/tetreault-2001-corpus-based-evaluation-centering-pronoun/) · [**2000** 1 work](https://md.research.elamur.ai/free/en/papers/stent-2000-rhetorical-structure-dialog/)

### 1990s

[**1999** 1 work](https://md.research.elamur.ai/free/en/papers/tetreault-1999-analysis-syntax-based-pronoun-resolution/)

### 1980s

[**1989** 1 work](https://md.research.elamur.ai/free/en/papers/hinkelman-1989-two-constraints-speech-act-ambiguity/)

### 1970s

[**1979** 1 work](https://md.research.elamur.ai/free/en/papers/allen-1979-plans-inference-indirect-speech-acts/)

## Concepts, shelves and topics

Of its works in our library: the number of works on each and its rating among them.

💡 Concepts 2[Semantic role labeling · 1 work](https://md.research.elamur.ai/free/en/concepts/semantic-role-labeling/) · [Context-free grammar · 1 work](https://md.research.elamur.ai/free/en/concepts/context-free-grammar/)

📚 Shelves 8[Natural language processing · 12 works](https://md.research.elamur.ai/free/en/shelves/nlp/) · [Linguistics & languages · 17 works](https://md.research.elamur.ai/free/en/shelves/linguistics/) · [AI & machine learning · 4 works](https://md.research.elamur.ai/free/en/shelves/ai/) · [English · 6 works](https://md.research.elamur.ai/free/en/shelves/en/) · [Machine translation · 5 works](https://md.research.elamur.ai/free/en/shelves/machine-translation/) · [Mathematics & statistics · 1 work](https://md.research.elamur.ai/free/en/shelves/mathematics/) · [Computing & GPUs · 1 work](https://md.research.elamur.ai/free/en/shelves/computing/) · [Machine translation · 2 works](https://md.research.elamur.ai/free/en/shelves/mtrans/)

🧭 Topics 18[Semantic role labeling · 1 work](https://md.research.elamur.ai/free/en/topics/semantic-role-labeling/) · [Evaluation & benchmarks · 5 works](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/) · [Vision & language · 1 work](https://md.research.elamur.ai/free/en/topics/multimodal/) · [Semantic parsing & meaning representation · 4 works](https://md.research.elamur.ai/free/en/topics/semantic-parsing/) · [Machine translation · 3 works](https://md.research.elamur.ai/free/en/topics/machine-translation/) · [Text generation · 2 works](https://md.research.elamur.ai/free/en/topics/text-generation/) · [Sentiment & emotion · 1 work](https://md.research.elamur.ai/free/en/topics/sentiment-emotion/) · [Named entity recognition · 1 work](https://md.research.elamur.ai/free/en/topics/named-entity-recognition/) · [Reasoning · 2 works](https://md.research.elamur.ai/free/en/topics/reasoning/) · [Speech act theory · 2 works](https://md.research.elamur.ai/free/en/topics/speech-act-theory/) · [Coreference resolution · 2 works](https://md.research.elamur.ai/free/en/topics/coreference-resolution/) · [Syntax & grammar formalisms · 2 works](https://md.research.elamur.ai/free/en/topics/syntax-grammar/) · [Language models & pretraining · 1 work](https://md.research.elamur.ai/free/en/topics/language-models/) · [Low-resource languages · 1 work](https://md.research.elamur.ai/free/en/topics/low-resource-languages/) · [Universal Dependencies · 2 works](https://md.research.elamur.ai/free/en/topics/universal-dependencies/)+3 more

Agglomerations: the GHSL Urban Centre Database 2025; populations: the GHSL functional urban areas 2015 (urban centre and commuting zone), else the urban centre, as each page says (European Commission, Joint Research Centre, CC BY 4.0). An institution belongs to the urban centre its coordinates (ROR, OpenAlex) fall in, or one within 5 km. Land: Natural Earth.

- [← back to](https://md.research.elamur.ai/free/en/countries/united-states/) — United States

Updated 23 October 2026