The overall structure of SDMAE.
<p>In sentence-aspect pairs encoder, mainly a BERT-based encoding process. DualGCN modules contain SemGCN and SynGCN. <i>Loss</i><sub><i>ar</i></sub> and <i>Loss</i><sub><i>scl</i></sub> refer to the multi-strategy auxilia...
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| مؤلفون آخرون: | , , , |
| منشور في: |
2025
|
| الموضوعات: | |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| _version_ | 1852017669901910016 |
|---|---|
| author | Lu Liu (171341) |
| author2 | Da Li (194457) Chuanxu Yue (22041810) Xiaojin Gao (11264337) Yunhai Zhu (19467015) |
| author2_role | author author author author |
| author_facet | Lu Liu (171341) Da Li (194457) Chuanxu Yue (22041810) Xiaojin Gao (11264337) Yunhai Zhu (19467015) |
| author_role | author |
| dc.creator.none.fl_str_mv | Lu Liu (171341) Da Li (194457) Chuanxu Yue (22041810) Xiaojin Gao (11264337) Yunhai Zhu (19467015) |
| dc.date.none.fl_str_mv | 2025-08-12T17:38:34Z |
| dc.identifier.none.fl_str_mv | 10.1371/journal.pone.0329018.g002 |
| dc.relation.none.fl_str_mv | https://figshare.com/articles/figure/The_overall_structure_of_SDMAE_/29895024 |
| dc.rights.none.fl_str_mv | CC BY 4.0 info:eu-repo/semantics/openAccess |
| dc.subject.none.fl_str_mv | Neuroscience Science Policy Mental Health Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Information Systems not elsewhere classified supervised contrastive learning strategy auxiliary enhancement model &# 8217 head attention mechanism feature aggregation system comprehensive experiments conducted art methods across sentiment polarity associated incorporates sentiment lexicons based sentiment analysis accurately detect sentiment original dependency tree xlink "> aspect aspect sentiment recognition extract syntactic features speech features dependency trees dependency parsing aspect terms syntactic structures syntactic denoising specific part significantly impairs several state semantic information public datasets primarily focused online reviews often insufficient often derived noise introduced maam ), integrate semantic existing studies context words constructing graphs adversely impacted |
| dc.title.none.fl_str_mv | The overall structure of SDMAE. |
| dc.type.none.fl_str_mv | Image Figure info:eu-repo/semantics/publishedVersion image |
| description | <p>In sentence-aspect pairs encoder, mainly a BERT-based encoding process. DualGCN modules contain SemGCN and SynGCN. <i>Loss</i><sub><i>ar</i></sub> and <i>Loss</i><sub><i>scl</i></sub> refer to the multi-strategy auxiliary enhancement loss.</p> |
| eu_rights_str_mv | openAccess |
| id | Manara_2f4f80b2fbe609376fd773bef42e59f7 |
| identifier_str_mv | 10.1371/journal.pone.0329018.g002 |
| network_acronym_str | Manara |
| network_name_str | ManaraRepo |
| oai_identifier_str | oai:figshare.com:article/29895024 |
| publishDate | 2025 |
| repository.mail.fl_str_mv | |
| repository.name.fl_str_mv | |
| repository_id_str | |
| rights_invalid_str_mv | CC BY 4.0 |
| spelling | The overall structure of SDMAE.Lu Liu (171341)Da Li (194457)Chuanxu Yue (22041810)Xiaojin Gao (11264337)Yunhai Zhu (19467015)NeuroscienceScience PolicyMental HealthEnvironmental Sciences not elsewhere classifiedBiological Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedsupervised contrastive learningstrategy auxiliary enhancementmodel &# 8217head attention mechanismfeature aggregation systemcomprehensive experiments conductedart methods acrosssentiment polarity associatedincorporates sentiment lexiconsbased sentiment analysisaccurately detect sentimentoriginal dependency treexlink "> aspectaspect sentiment recognitionextract syntactic featuresspeech featuresdependency treesdependency parsingaspect termssyntactic structuressyntactic denoisingspecific partsignificantly impairsseveral statesemantic informationpublic datasetsprimarily focusedonline reviewsoften insufficientoften derivednoise introducedmaam ),integrate semanticexisting studiescontext wordsconstructing graphsadversely impacted<p>In sentence-aspect pairs encoder, mainly a BERT-based encoding process. DualGCN modules contain SemGCN and SynGCN. <i>Loss</i><sub><i>ar</i></sub> and <i>Loss</i><sub><i>scl</i></sub> refer to the multi-strategy auxiliary enhancement loss.</p>2025-08-12T17:38:34ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0329018.g002https://figshare.com/articles/figure/The_overall_structure_of_SDMAE_/29895024CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/298950242025-08-12T17:38:34Z |
| spellingShingle | The overall structure of SDMAE. Lu Liu (171341) Neuroscience Science Policy Mental Health Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Information Systems not elsewhere classified supervised contrastive learning strategy auxiliary enhancement model &# 8217 head attention mechanism feature aggregation system comprehensive experiments conducted art methods across sentiment polarity associated incorporates sentiment lexicons based sentiment analysis accurately detect sentiment original dependency tree xlink "> aspect aspect sentiment recognition extract syntactic features speech features dependency trees dependency parsing aspect terms syntactic structures syntactic denoising specific part significantly impairs several state semantic information public datasets primarily focused online reviews often insufficient often derived noise introduced maam ), integrate semantic existing studies context words constructing graphs adversely impacted |
| status_str | publishedVersion |
| title | The overall structure of SDMAE. |
| title_full | The overall structure of SDMAE. |
| title_fullStr | The overall structure of SDMAE. |
| title_full_unstemmed | The overall structure of SDMAE. |
| title_short | The overall structure of SDMAE. |
| title_sort | The overall structure of SDMAE. |
| topic | Neuroscience Science Policy Mental Health Environmental Sciences not elsewhere classified Biological Sciences not elsewhere classified Information Systems not elsewhere classified supervised contrastive learning strategy auxiliary enhancement model &# 8217 head attention mechanism feature aggregation system comprehensive experiments conducted art methods across sentiment polarity associated incorporates sentiment lexicons based sentiment analysis accurately detect sentiment original dependency tree xlink "> aspect aspect sentiment recognition extract syntactic features speech features dependency trees dependency parsing aspect terms syntactic structures syntactic denoising specific part significantly impairs several state semantic information public datasets primarily focused online reviews often insufficient often derived noise introduced maam ), integrate semantic existing studies context words constructing graphs adversely impacted |