five

Italian Verb Lexicon for Sentiment Inference

收藏
NIAID Data Ecosystem2026-05-02 收录
下载链接:
https://zenodo.org/record/8195008
下载链接
链接失效反馈
官方服务:
资源简介:
Italian Verb Lexicon for Sentiment Inference Theory: For a description of the theory behind the specifications of the corpus, please read the attached paper.  Example of json entry: {"verb": "soddisfare", "frames": [{"fillers": ["Subj", "DirObj/IndObj"], "polarity": "POS", "effects": [["DirObj/IndObj", "pos"]], "expectations": [], "examples": ["L'offerta ha soddisfatto i clienti.", "Soddisfare al pubblico."], "remarks": [], "relations": [["Subj", "DirObj/IndObj", "pro"]]}]} Description:  The verb "soddisfare" has 2 frames, a subject followed by a direct or indirect object. The verb polarity is positive. There is an positive effect on the direct (indirect) object. No expectations. There is a in favour (pro) relation from the subject to the direct (indirect) object. Two example sentenes are given. Synonyms: Some entries are references to synonym verbs with identical frames: {"verb": "consacrare", "germanTranslation": "widmen", "frameReference": "dedicare", "examples": ["Consacrare tempo alle sue passioni"]} Here, "cosacrare" and "dedicare" are assumed synonyms with the same syntactic frames. Used Tags: A few explanations on the tags used in the verb specifications sheets. Subj = subject DirObj = direct object, as in "Il professore legge __il giornale__". IndObj = indirect object, as in "Permettere qualcosa __a qualcuno__". RefObj = reflexive object (pronoun), as in "La squadra avversaria __si__ è arrabbiata moltissimo".  PrepObj[prep] = prepositional phrase; the preposition is specified in the square brackets. If more than one preposition can occur,                 no specification is given. SubCl = a subordinate clause, usually introduced by "che" or "di" such as in "Ha detto __di andarsene__",                  "Ha detto __che tutto è andato bene__". mod = any type of modifier, mostly adverbs, e.g. "Se ne è andato __subito__". *, e.g. mod* = indicates optionality
创建时间:
2024-07-11
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作