About

Ralph Grishman is a professor in the Department of Computer Science at New York University, where he has served since 1983 and was chairman from 1986 to 1988. His academic background includes a Ph.D. in Physics from Columbia University, earned in 1973, with a thesis on Numerical Studies of Self-Avoiding Walks, and an A.B. in Physics from Columbia College, graduated summa cum laude in 1968. His early career involved roles such as Assistant Professor, Associate Professor, and Associate Research Scientist at the Courant Institute of Mathematical Sciences, NYU, and he has also held positions at Barnard College and Columbia University, among others. His professional service includes leadership roles in the Association for Computational Linguistics, where he served as Vice President and President, and participation in government committees such as ARPA and NIST, focusing on speech, natural language processing, and text analysis. His research contributions are centered on natural language processing, computational linguistics, sublanguage processing, and language analysis, with a significant focus on developing computational models for language understanding, information extraction, and domain-specific language analysis.

Research topics

• Artificial Intelligence
• Machine Learning
• Computer Science
• Data Mining
• Natural Language Processing
• Information Retrieval
• Mathematics
• Programming language

Selected publications

• MUCking In, or Fifty Years in Information Extraction

  Computational Linguistics · 2025-01-01

  articleOpen access1st authorCorresponding

  Abstract I want to thank the ACL for this Lifetime Achievement Award. I am deeply honored to be receiving it. I would also like to thank the students, faculty, and researchers who were members of the Proteus Project during most of my professional lifetime. It was an honor to serve that group.

  PublisherDOI

• Lexicalized Dependency Paths Based Supervised Learning for Relation Extraction

  Computer Systems Science and Engineering · 2022 · 85 citations

  Senior authorCorresponding
  • Computer Science
  • Computer Science
  • Artificial Intelligence

  Log-linear models and more recently neural network models used for supervised relation extraction requires substantial amounts of training data and time, limiting the portability to new relations and domains. To this end, we propose a training representation based on the dependency paths between entities in a dependency tree which we call lexicalized dependency paths (LDPs). We show that this representation is fast, efficient and transparent. We further propose representations utilizing entity types and its subtypes to refine our model and alleviate the data sparsity problem. We apply lexicalized dependency paths to supervised learning using the ACE corpus and show that it can achieve similar performance level to other state-of-the-art methods and even surpass them on several categories.

  PublisherOA PDFDOI

• Employing Lexicalized Dependency Paths for Active Learning of Relation Extraction

  Intelligent Automation & Soft Computing · 2022 · 53 citations

  Senior authorCorresponding
  • Computer Science
  • Computer Science
  • Artificial Intelligence

  Active learning methods which present selected examples from the <i>corpus</i> for annotation provide more efficient learning of supervised relation extraction models, but they leave the developer in the unenviable role of a passive informant. To restore the developer’s proper role as a partner with the system, we must give the developer an ability to inspect the extraction model during development. We propose to make this possible through a representation based on lexicalized dependency paths (LDPs) coupled with an active learner for LDPs. We apply LDPs to both simulated and real active learning with ACE as evaluation and a year’s newswire for training and show that simulated active learning greatly reduces annotation cost while maintaining similar performance level of supervised learning, while real active learning yields comparable performance to the state-of-the-art using a small number of annotations.

  PublisherOA PDFDOI

• Learning Relatedness between Types with Prototypes for Relation Extraction

  2021 · 7 citations

  Senior authorCorresponding
  • Computer Science
  • Computer Science
  • Artificial Intelligence

  Relation schemas are often pre-defined for each relation dataset. Relation types can be related from different datasets and have overlapping semantics. We hypothesize we can combine these datasets according to the semantic relatedness between the relation types to overcome the problem of lack of training data. It is often easy to discover the connection between relation types based on relation names or annotation guides, but hard to measure the exact similarity and take advantage of the connection between the relation types from different datasets. We propose to use prototypical examples to represent each relation type and use these examples to augment related types from a different dataset. We obtain further improvement (ACE05) with this type augmentation over a strong baseline which uses multi-task learning between datasets to obtain better feature representation for relations. We make our implementation publicly available:

  PublisherOA PDFDOI

• Twenty-five years of information extraction

  Natural Language Engineering · 2019-09-20 · 62 citations

  article1st authorCorresponding

  Abstract Information extraction is the process of converting unstructured text into a structured data base containing selected information from the text. It is an essential step in making the information content of the text usable for further processing. In this paper, we describe how information extraction has changed over the past 25 years, moving from hand-coded rules to neural networks, with a few stops on the way. We connect these changes to research advances in NLP and to the evaluations organized by the US Government.

  PublisherDOI

• Including New Patterns to Improve Event Extraction Systems

  2018-05-10 · 7 citations

  articleSenior author

  Event Extraction (EE) is a challenging Information Extraction task which aims to discover event triggers of specific types along with their arguments. Most recent research on Event Extraction relies on pattern-based or feature-based approaches, trained on annotated corpora, to recognize combi- nations of event triggers, arguments, and other contextual in- formation. However, as the event instances in the ACE corpus are not evenly distributed, some frequent expressions involving ACE event triggers do not appear in the training data, adversely affecting the performance. In this paper, we demon- strate the effectiveness of systematically importing expert-level patterns from TABARI to boost EE performance. The experimental results demonstrate that our pattern-based sys- tem with the expanded patterns can achieve 69.8% (with 1.9% absolute improvement) F-measure over the baseline, an advance over current state-of-the-art systems.

  Publisher

• A Survey of Syntactic Analysis Procedures for Natural Language

  Defense Technical Information Center (DTIC) · 2018-03-03 · 16 citations

  book1st authorCorresponding

  The report includes a brief discussion of the role of automatic syntactic analysis, a survey of parsing procedures, past and present, and a discussion of the approaches taken to a number of difficult linguistic problems, such as conjunction and graded acceptability. It also contains precise specifications in the programming language SETL of a number of parsing algorithms.

  Publisher

• Graph Convolutional Networks With Argument-Aware Pooling for Event Detection

  Proceedings of the AAAI Conference on Artificial Intelligence · 2018-04-26 · 382 citations

  articleOpen accessSenior author

  The current neural network models for event detection have only considered the sequential representation of sentences. Syntactic representations have not been explored in this area although they provide an effective mechanism to directly link words to their informative context for event detection in the sentences. In this work, we investigate a convolutional neural network based on dependency trees to perform event detection. We propose a novel pooling method that relies on entity mentions to aggregate the convolution vectors. The extensive experiments demonstrate the benefits of the dependency-based convolutional neural networks and the entity mention-based pooling method for event detection. We achieve the state-of-the-art performance on widely used datasets with both perfect and predicted entity mentions.

  PublisherOA PDFDOI

• A Case Study on Learning a Unified Encoder of Relations

  2018-01-01 · 4 citations

  articleOpen accessSenior author

  Typical relation extraction models are trained on a single corpus annotated with a pre-defined relation schema. An individual corpus is often small, and the models may often be biased or overfitted to the corpus. We hypothesize that we can learn a better representation by combining multiple relation datasets. We attempt to use a shared encoder to learn the unified feature representation and to augment it with regularization by adversarial training. The additional corpora feeding the encoder can help to learn a better feature representation layer even though the relation schemas are different. We use ACE05 and ERE datasets as our case study for experiments. The multi-task model obtains significant improvement on both datasets.

  PublisherOA PDFDOI

• Active learning based named entity recognition and its application in natural language coverless information hiding

  網際網路技術學刊 · 2017-03-01 · 14 citations

  article

  Named entity recognition systems trained on one domain usually have a substantial drop in performance when applied to a different domain. In this paper, we apply active learning to domain adaptation for named entity recognition systems, propose various sampling optimizations, and show that the labeling effort can be reduced by over 92% while achieving the same performance as supervised method. Named entity recognition can be effectively applied to information extraction, machine translation, text classification and many other areas. We propose a new application area for named entity recognition, namely in natural language information hiding: a novel coverless information hiding method based on text big data is proposed, utilizing named entities to mark the locations of the hidden information. Coverless information hiding is a brand new area of information hiding that achieves the transmission of hidden information without any modification in the carrier text. Furthermore, active learning allows our information hiding method to be applied to text from new domains without substantial labeling effort.

  Publisher

Frequent coauthors

• Adam Meyers

  40 shared

• Catherine Macleod

  Bangor University

  35 shared

• Heng Ji

  25 shared

• Roman Yangarber

  24 shared

• John Sterling

  21 shared

• Satoshi Sekine

  18 shared

• Lynette Hirschman

  Mitre (United States)

  17 shared

• Bonan Min

  16 shared

Labs

• NYU Courant Institute of Mathematical SciencesPI