Jseg

A modified version of Jieba

Equipped with Emoticon detection
Emoticons will not be segmented as sequences of meaningless punctuations.
Data are trained with Sinica Corpus
Results are more accurate when dealing with Traditional Chinese (F1-score = 0.91).
Using Brill Tagger
Training data are trained with Sinica Treebank, which raises the accuracy of POS tagging.

pip install -U jseg

from jseg import Jieba
j = Jieba()

j.add_guaranteed_wordlist(lst)

Here's a sample text:

sample = '期末要爆炸啦！ ◢▆▅▄▃崩╰(〒皿〒)╯潰▃▄▅▇◣'

Segmentation with POS (part-of-speech)

j.seg(sample, pos=True)