Search

Your search keyword '"Hockenmaier, Julia"' showing total 29 results
Start Over Author "Hockenmaier, Julia" Publisher association for computational linguistics
29 results on '"Hockenmaier, Julia"'

10. Towards Problem Solving Agents that Communicate and Learn

Author

Narayan-Chen, Anjali, primary, Graber, Colin, additional, Das, Mayukh, additional, Islam, Md Rakibul, additional, Dan, Soham, additional, Natarajan, Sriraam, additional, Doppa, Janardhan Rao, additional, Hockenmaier, Julia, additional, Palmer, Martha, additional, and Roth, Dan, additional

Published
2017
Full Text
View/download PDF

24. Example selection for bootstrapping statistical parsers

Author

Steedman, Mark, primary, Hwa, Rebecca, additional, Clark, Stephen, additional, Osborne, Miles, additional, Sarkar, Anoop, additional, Hockenmaier, Julia, additional, Ruhlen, Paul, additional, Baker, Steven, additional, and Crim, Jeremiah, additional

Published
2003
Full Text
View/download PDF

25. Bootstrapping statistical parsers from small datasets

Author

Steedman, Mark, primary, Osborne, Miles, additional, Sarkar, Anoop, additional, Clark, Stephen, additional, Hwa, Rebecca, additional, Hockenmaier, Julia, additional, Ruhlen, Paul, additional, Baker, Steven, additional, and Crim, Jeremiah, additional

Published
2003
Full Text
View/download PDF

28. An Encoder-Decoder Approach to the Paradigm Cell Filling Problem

Author

Miikka Silfverberg, Mans Hulden, Riloff, Ellen, Chiang , David, Hockenmaier, Julia, Jun’ichi , Tsujii, Department of Digital Humanities, Language Technology, University Management, and Faculty of Arts

Subjects
060201 languages & linguistics, Theoretical computer science, Training set, Computer science, education, 06 humanities and the arts, 010501 environmental sciences, 113 Computer and information sciences, 01 natural sciences, Task (project management), 0602 languages and literature, Inflection, Code (cryptography), 6121 Languages, Encoder decoder, Filling Problem, Word (computer architecture), 0105 earth and related environmental sciences
Abstract

The Paradigm Cell Filling Problem in morphology asks to complete word inflection tables from partial ones. We implement novel neural models for this task, evaluating them on 18 data sets in 8 languages, showing performance that is comparable with previous work with far less training data. We also publish a new dataset for this task and code implementing the system described in this paper.

Published
2018

29. Adversarial Propagation and Zero-Shot Cross-Lingual Transfer of Word Vector Specialization

Author

Goran Glavaš, Ivan Vulić, Edoardo Maria Ponti, Anna Korhonen, Nikola Mrkšić, Riloff, Ellen, Chiang, David, Hockenmaier, Julia, and Tsujii, Jun'ichi

Subjects
FOS: Computer and information sciences, Vocabulary, Computer Science - Computation and Language, Lexical simplification, Computer science, business.industry, media_common.quotation_subject, WordNet, 02 engineering and technology, 010501 environmental sciences, computer.software_genre, 01 natural sciences, Adversarial system, 0202 electrical engineering, electronic engineering, information engineering, 020201 artificial intelligence & image processing, Artificial intelligence, Dialog box, business, Computation and Language (cs.CL), computer, Natural language processing, 0105 earth and related environmental sciences, media_common
Abstract

Semantic specialization is the process of fine-tuning pre-trained distributional word vectors using external lexical knowledge (e.g., WordNet) to accentuate a particular semantic relation in the specialized vector space. While post-processing specialization methods are applicable to arbitrary distributional vectors, they are limited to updating only the vectors of words occurring in external lexicons (i.e., seen words), leaving the vectors of all other words unchanged. We propose a novel approach to specializing the full distributional vocabulary. Our adversarial post-specialization method propagates the external lexical knowledge to the full distributional space. We exploit words seen in the resources as training examples for learning a global specialization function. This function is learned by combining a standard L2-distance loss with an adversarial loss: the adversarial component produces more realistic output vectors. We show the effectiveness and robustness of the proposed method across three languages and on three tasks: word similarity, dialog state tracking, and lexical simplification. We report consistent improvements over distributional word vectors and vectors specialized by other state-of-the-art specialization frameworks. Finally, we also propose a cross-lingual transfer method for zero-shot specialization which successfully specializes a full target distributional space without any lexical knowledge in the target language and without any bilingual data., Accepted at EMNLP 2018

Published
2018

Catalog

Books, media, physical & digital resources
See catalog results