Which hyper parameter mostly influence the number of topics for Chinese texts?

Nobody has claimed this yet.

```
import jieba
def tokenize_zh(text):
words = jieba.lcut(text)
words = list(filter(lambda x: (len(x)>1), words))
return words

import numpy as np
from umap import UMAP
from sklearn.feature_extraction.text import CountVectorizer
from sentence_transformers import SentenceTransformer
embedding_model = SentenceTransformer("BAAI/bge-base-zh-v1.5") # a Chinese embedding model
vectorizer_model = CountVectorizer(tokenizer=tokenize_zh, stop_words=stopwords, ngram_range=(1, 3), min_df=3) # The variiable stopwords is a list.
umap_model = UMAP(n_neighbors=15, n_components=5, min_dist=0.0, metric='cosine', random_state=42)
topic_model = BERTopic(language='chinese', embedding_model=embedding_model, umap_model=umap_model, top_n_words=10, min_topic_size=10, n_gram_range=(1,3), vectorizer_model=vectorizer_model, calculate_probabilities=True, verbose=True)

topics, probs = topic_model.fit_transform(docs)
topic_model.get_topic_info()
```

It has only 3 topics as below, much less than the Mallet tool.

I almost test every hyper parameters here, at last found the `n_neighbors `in the `UMAP `function works most apparently, however, even changing it from 15 to 50 only adds 1 new topic as below:

It seems BERTopic need more tuning parameters for Chinese or even CJK texts. Can anyone share some experience, please?

Start with the supplied Chinese example and inspect how BERTopic applies the UMAP, CountVectorizer, and topic-size settings during fit_transform. Reproduce the topic-count changes while varying the named parameters; done would require a documented, reproducible explanation or a clearly scoped improvement for Chinese or CJK behavior.

Written by the indexing model from the issue text.