Filters








372 Hits in 6.4 sec

Better Document-level Sentiment Analysis from RST Discourse Parsing [article]

Parminder Bhatia and Yangfeng Ji and Jacob Eisenstein
2015 arXiv   pre-print
We show that the discourse analyses produced by Rhetorical Structure Theory (RST) parsers can improve document-level sentiment analysis, via composition of local information up the discourse tree.  ...  Discourse structure is the hidden link between surface features and document-level properties, such as sentiment polarity.  ...  The time is therefore right to reconsider the effectiveness of RST for document-level sentiment analysis.  ... 
arXiv:1509.01599v2 fatcat:b5jfxo47sjdsjkfjtfordourui

Better Document-level Sentiment Analysis from RST Discourse Parsing

Parminder Bhatia, Yangfeng Ji, Jacob Eisenstein
2015 Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing  
We show that the discourse analyses produced by Rhetorical Structure Theory (RST) parsers can improve document-level sentiment analysis, via composition of local information up the discourse tree.  ...  Discourse structure is the hidden link between surface features and document-level properties, such as sentiment polarity.  ...  The time is therefore right to reconsider the effectiveness of RST for document-level sentiment analysis.  ... 
doi:10.18653/v1/d15-1263 dblp:conf/emnlp/BhatiaJE15 fatcat:ahst7n7cybfyrjknl6myfgrcgi

Predicting Discourse Structure using Distant Supervision from Sentiment [article]

Patrick Huber, Giuseppe Carenini
2019 arXiv   pre-print
We propose a novel approach that uses distant supervision on an auxiliary task (sentiment classification), to generate abundant data for RST-style discourse structure prediction.  ...  Our approach combines a neural variant of multiple-instance learning, using document-level supervision, with an optimal CKY-style tree generation algorithm.  ...  Our assumption is that such synergies between sentiment analysis and discourse parsing are bidirectional.  ... 
arXiv:1910.14176v1 fatcat:ulcn6rvgtrhlnb7zfddmezg2qu

Predicting Discourse Structure using Distant Supervision from Sentiment

Patrick Huber, Giuseppe Carenini
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
Our approach combines a neural variant of multipleinstance learning, using document-level supervision, with an optimal CKY-style tree generation algorithm.  ...  Discourse parsing could not yet take full advantage of the neural NLP revolution, mostly due to the lack of annotated datasets.  ...  Our assumption is that such synergies between sentiment analysis and discourse parsing are bidirectional.  ... 
doi:10.18653/v1/d19-1235 dblp:conf/emnlp/HuberC19 fatcat:asztbqmtgjd5nmbmoaci4nnp4y

Exploring Joint Neural Model for Sentence Level Discourse Parsing and Sentiment Analysis

Bita Nejat, Giuseppe Carenini, Raymond Ng
2017 Proceedings of the 18th Annual SIGdial Meeting on Discourse and Dialogue  
Discourse Parsing and Sentiment Analysis are two fundamental tasks in Natural Language Processing that have been shown to be mutually beneficial.  ...  Next, we apply three different Recursive Neural Net models: one for discourse structure prediction, one for discourse relation prediction and one for sentiment analysis.  ...  Sentiment at document level.  ... 
doi:10.18653/v1/w17-5535 dblp:conf/sigdial/NejatCN17 fatcat:khab45b5mzfu7grcxej2lpoq54

From Sentiment Annotations to Sentiment Prediction through Discourse Augmentation [article]

Patrick Huber, Giuseppe Carenini
2020 arXiv   pre-print
Sentiment analysis, especially for long documents, plausibly requires methods capturing complex linguistics structures.  ...  To accommodate this, we propose a novel framework to exploit task-related discourse for the task of sentiment analysis.  ...  Related Work This work is located at the intersection of recent approaches on discourse parsing and sentiment analysis and mostly influenced by four lines of research: (1) RST-style Discourse Parsing is  ... 
arXiv:2011.03021v1 fatcat:tcwuscgrmzhthik7sxwfuxc7tq

Predicting Above-Sentence Discourse Structure Using Distant Supervision from Topic Segmentation

Patrick Huber, Linzi Xing, Giuseppe Carenini
2022 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
RST-style discourse parsing plays a vital role in many NLP tasks, revealing the underlying semantic/pragmatic structure of potentially complex and diverse documents.  ...  To overcome the data sparsity issue, distantly supervised approaches from tasks like sentiment analysis and summarization have been recently proposed.  ...  , generally evaluating complete discourse trees from EDU-to-document level.  ... 
doi:10.1609/aaai.v36i10.21325 fatcat:ldgsnyoswfhgpouic5euyau7fy

DMRST: A Joint Framework for Document-Level Multilingual RST Discourse Segmentation and Parsing [article]

Zhengyuan Liu, Ke Shi, Nancy F. Chen
2021 arXiv   pre-print
In this work, we propose a document-level multilingual RST discourse parsing framework, which conducts EDU segmentation and discourse tree parsing jointly.  ...  Experimental results show that our model achieves state-of-the-art performance on document-level multilingual RST parsing in all sub-tasks.  ...  Acknowledgments This research was supported by funding from the Institute for Infocomm Research (I2R) under A*STAR ARES, Singapore.  ... 
arXiv:2110.04518v1 fatcat:jl4ugdztdbe77gxxi63q4dhbja

Evaluating Discourse in Structured Text Representations [article]

Elisa Ferracane, Greg Durrett, Junyi Jessy Li, Katrin Erk
2019 arXiv   pre-print
capturing discourse structure when compared to discourse dependency trees from an existing discourse parser.  ...  Liu and Lapata (2018) propose a structured attention mechanism for text classification that derives a tree over a text, akin to an RST discourse tree.  ...  The first author was supported by the NSF Graduate Research Fellowship Program under Grant No. 2017247409. 9 More than 4 levels caused training to become unstable. 10 Less than 25% of trees in the RST  ... 
arXiv:1906.01472v2 fatcat:lb2iw3jbzfcatpfxkve3wvmrwy

Evaluating Discourse in Structured Text Representations

Elisa Ferracane, Greg Durrett, Junyi Jessy Li, Katrin Erk
2019 Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics  
capturing discourse structure when compared to discourse dependency trees from an existing discourse parser.  ...  Liu and Lapata (2018) propose a structured attention mechanism for text classification that derives a tree over a text, akin to an RST discourse tree.  ...  ., 2017) , and sentiment analysis (Ji and Smith, 2017) .  ... 
doi:10.18653/v1/p19-1062 dblp:conf/acl/FerracaneDLE19 fatcat:yxaeq2jhkjestcqbqvtxsnbvnu

W-RST: Towards a Weighted RST-style Discourse Framework [article]

Patrick Huber, Wen Xiao, Giuseppe Carenini
2021 arXiv   pre-print
In particular, we find that weighted discourse trees from auxiliary tasks can benefit key NLP downstream applications, compared to nuclearity-centered approaches.  ...  Aiming for a better integration of data-driven and linguistically-inspired approaches, we explore whether RST Nuclearity, assigning a binary assessment of importance between text segments, can be replaced  ...  attributions can be generated from W-RST-Sent and W-RST-Summ and used for the sentiment analysis and summarization downstream tasks.  ... 
arXiv:2106.02658v1 fatcat:x4qbguvj2vfpvcu3nebcfdt2la

Where Are We in Discourse Relation Recognition?

Katherine Atwell, Junyi Jessy Li, Malihe Alikhani
2021 SIGDIAL Conferences  
We present a position paper which provides a systematic analysis of the state of the art discourse parsers.  ...  However it is often difficult to achieve good results from current discourse models, largely due to the difficulty of the task, particularly recognizing implicit discourse relations.  ...  Introduction Discourse analysis is a crucial analytic level in NLP.  ... 
dblp:conf/sigdial/AtwellLA21 fatcat:6lee2gainjasdioejdyb3idarm

Neural Discourse Structure for Text Categorization

Yangfeng Ji, Noah A. Smith
2017 Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)  
Our approach uses a recursive neural network and a newly proposed attention mechanism to compute a representation of the text that focuses on salient content, from the perspective of both RST and the task  ...  We show that discourse structure, as defined by Rhetorical Structure Theory and provided by an existing discourse parser, benefits text categorization.  ...  Discourse structure was not considered. Discourse for sentiment analysis.  ... 
doi:10.18653/v1/p17-1092 dblp:conf/acl/JiS17 fatcat:kji5xpzccfaffkcsphzybbzqim

Neural Discourse Structure for Text Categorization [article]

Yangfeng Ji, Noah Smith
2017 arXiv   pre-print
Our approach uses a recursive neural network and a newly proposed attention mechanism to compute a representation of the text that focuses on salient content, from the perspective of both RST and the task  ...  We show that discourse structure, as defined by Rhetorical Structure Theory and provided by an existing discourse parser, benefits text categorization.  ...  Discourse structure was not considered. Discourse for sentiment analysis.  ... 
arXiv:1702.01829v2 fatcat:ygfhyhpuknfjhnntmpuryiwfwy

Sentiment analysis based on rhetorical structure theory: Learning deep neural networks from discourse trees [article]

Mathias Kraus, Stefan Feuerriegel
2017 arXiv   pre-print
As a remedy, we develop a discourse-aware method that builds upon the discourse structure of documents.  ...  Prominent applications of sentiment analysis are countless, covering areas such as marketing, customer service and communication.  ...  Section 2 reviews discourse parsing and RST-based sentiment analysis. Section 3 then introduces our Discourse-LSTM, as well as our algorithms for data augmentation.  ... 
arXiv:1704.05228v2 fatcat:ujty7527dvb6dcipywslqeoq2a
« Previous Showing results 1 — 15 out of 372 results