{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/multi-task-learning-for-target-dependent","title":"Multi-task Learning for Target-dependent Sentiment Classification","arxiv_id":"1902.02930","date":"2019-02-08","proceeding":null,"authors":["Divam Gupta","Kushagra Singh","Soumen Chakrabarti","Tanmoy Chakraborty"],"abstract":"Detecting and aggregating sentiments toward people, organizations, and events\nexpressed in unstructured social media have become critical text mining\noperations. Early systems detected sentiments over whole passages, whereas more\nrecently, target-specific sentiments have been of greater interest. In this\npaper, we present MTTDSC, a multi-task target-dependent sentiment\nclassification system that is informed by feature representation learnt for the\nrelated auxiliary task of passage-level sentiment classification. The auxiliary\ntask uses a gated recurrent unit (GRU) and pools GRU states, followed by an\nauxiliary fully-connected layer that outputs passage-level predictions. In the\nmain task, these GRUs contribute auxiliary per-token representations over and\nabove word embeddings. The main task has its own, separate GRUs. The auxiliary\nand main GRUs send their states to a different fully connected layer, trained\nfor the main task. Extensive experiments using two auxiliary datasets and three\nbenchmark datasets (of which one is new, introduced by us) for the main task\ndemonstrate that MTTDSC outperforms state-of-the-art baselines. Using\nword-level sensitivity analysis, we present anecdotal evidence that prior\nsystems can make incorrect target-specific predictions because they miss\nsentiments expressed by words independent of target.","url_abs":"http://arxiv.org/abs/1902.02930v1","url_pdf":"http://arxiv.org/pdf/1902.02930v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"multi-task-learning-for-target-dependent","repo_url":"https://github.com/divamgupta/mttdsc","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"multi-task-learning","task_name":"Multi-Task Learning"},{"task_slug":"sentiment-analysis","task_name":"Sentiment Analysis"},{"task_slug":"sentiment-classification","task_name":"Sentiment Classification"},{"task_slug":"word-embeddings","task_name":"Word Embeddings"}],"methods":[{"method_slug":"gru","method_name":"GRU"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}