{"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/-hot-lexicon-embedding-based-two-level-lstm","title":"$ρ$-hot Lexicon Embedding-based Two-level LSTM for Sentiment Analysis","arxiv_id":"1803.07771","date":"2018-03-21","proceeding":null,"authors":["Ou Wu","Tao Yang","Mengyang Li","Ming Li"],"abstract":"Sentiment analysis is a key component in various text mining applications.\nNumerous sentiment classification techniques, including conventional and deep\nlearning-based methods, have been proposed in the literature. In most existing\nmethods, a high-quality training set is assumed to be given. Nevertheless,\nconstructing a high-quality training set that consists of highly accurate\nlabels is challenging in real applications. This difficulty stems from the fact\nthat text samples usually contain complex sentiment representations, and their\nannotation is subjective. We address this challenge in this study by leveraging\na new labeling strategy and utilizing a two-level long short-term memory\nnetwork to construct a sentiment classifier. Lexical cues are useful for\nsentiment analysis, and they have been utilized in conventional studies. For\nexample, polar and privative words play important roles in sentiment analysis.\nA new encoding strategy, that is, $\\rho$-hot encoding, is proposed to alleviate\nthe drawbacks of one-hot encoding and thus effectively incorporate useful\nlexical cues. We compile three Chinese data sets on the basis of our label\nstrategy and proposed methodology. Experiments on the three data sets\ndemonstrate that the proposed method outperforms state-of-the-art algorithms.","url_abs":"http://arxiv.org/abs/1803.07771v1","url_pdf":"http://arxiv.org/pdf/1803.07771v1.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":"-hot-lexicon-embedding-based-two-level-lstm","repo_url":"https://github.com/Tju-AI/two-stage-labeling-for-the-sentiment-orientations","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"sentiment-analysis","task_name":"Sentiment Analysis"},{"task_slug":"sentiment-classification","task_name":"Sentiment Classification"},{"task_slug":"two","task_name":"Vocal Bursts Valence Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}