Abstract | ||
---|---|---|
For foreign Chinese learners, learning Pinyin is much easier than Chinese characters. Traditional Chinese input methods (IMEs) that are intended for Chinese people are not suitable for Chinese learners whose mother language is non-Chinese. In general, different languages can be translated into each other. By this feature and the relevant knowledge of existing IMEs, we take Chinese and English as an example to propose a hybrid encoding that combines Pinyin and the first two letters of its translation. We implement SeeIME with this hybrid Pinyin code. SeeIME can effectively reduce the number of candidate words and help users recognize Chinese characters better via a three-block display method. Using this code for sentence generation can improve the hit rate, which increases from 58% to 84%. The implemented IME is already available for the Android system.
|
Year | DOI | Venue |
---|---|---|
2019 | 10.1145/3321408.3321594 | Proceedings of the ACM Turing Celebration Conference - China |
Keywords | Field | DocType |
hybrid code, language model, soft keyboard | Pinyin,Computer science,Natural language processing,Artificial intelligence | Conference |
ISBN | Citations | PageRank |
978-1-4503-7158-2 | 0 | 0.34 |
References | Authors | |
0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jian Yao | 1 | 9 | 3.61 |
Xingfa Shen | 2 | 11 | 4.64 |
Jianhui Qiu | 3 | 0 | 0.34 |
Ting Wang | 4 | 725 | 120.28 |