NLU Datasets

The 850,000 English Japanese Parallel Corpus Data is a bilingual text is stored in text format. It covers multiple fields such as tourism, medical treatment, daily life, news, etc. average English sentence 23 words. The data desensitization and quality checking had been done. It can be used as a basic corpus for text data analysis in fields such as machine translation.

Japanese and English parallel corpus, 380,000 groups in total; excluded political, porn, personal information and other sensitive vocabulary; it can be a base corpus for text-based data analysis, used in machine translation and other fields.

9.83 Million Pairs of Sentences - Chinese-Japanese Parallel Corpus Data be stored in txt format. It covers multiple fields including general, IT, news, patent, and international engine. The data desensitization and quality checking had been done. It can be used as a basic corpus for text data analysis in fields such as machine translation.

10,000 Chinese news event annotated data. The contents are hot news in 2013. Each piece of news contains one or more events. Each event is annotated. The data is stored in xml and can be used for natural language understanding.

8,178 Chinese social comments annotated data. The contents are hot news in 2013. Each piece of news contains one or more events and is annotated with time, theme, cause, procedure and result. The data is stored in xml and can be used for natural language understanding.

It collectes comments from different car forums and fine-grained annotation is carried out on posts commented by users. Annotations include labels of manufacturer, brand, model, attribute, description value, tendency, etc. It can be used in fine-grained natural language understanding research, emotion analysis and some other fields.

Annotation of 687,694 sentences generated by users in the mobile phone scene, covering to-do scenes, location scenes, and schedule scenes. The data set can be used for natural language understanding tasks.

Intent-like single-sentence annotated textual data, the data size is 28,237 sentences, artificially written, and annotated with intent classes, including slot and slot value information; the intent field includes music, weather, date, schedule, home equipment, etc.; it is applied to intent recognition research and related fields.

Intent-like single-sentence annotated textual data, the data size is 47811 sentences, annotated with intent classes, including slot and slot value information; the intent field includes music, weather, date, schedule, home equipment, etc.; it is applied to intent recognition research and related fields.