the score of clus_center is smaller than filter_thre, so it doesn't occur in keywords.txt, leading to KeyError in embs[cate_ph]

[chanwing]

Dear authors of TaxoGen, The keyword score of clus_center is smaller than filter_thre, so it (the clus_center) doesn't occur in keywords.txt, leading to KeyError in embs[cate_ph]. How to solve this problem? 作者们您们好， 聚类中心的词的分数小于阈值，因此该词不出现在keywords.txt，导致embs词典里没有这个词，embs[cate_ph]时查不到该词，出现KeyError，请问如何解决？

[chanwing]

Traceback (most recent call last): File "main.py", line 140, in main(opt) File "main.py", line 117, in main recur(input_dir, root_dir, n_cluster, '*', n_cluster_iter, filter_thre, n_expand, level, True, True) File "main.py", line 103, in recur filter_thre, n_expand, level + 1, caseolap, local_embedding) File "main.py", line 103, in recur filter_thre, n_expand, level + 1, caseolap, local_embedding) File "main.py", line 97, in recur main_local_embedding(node_dir, df.doc_file, df.index_file, parent, n_expand) File "/Users/blabla/PycharmProjects/taxogenpy3/code/local_embedding_training.py", line 147, in main_local_embedding cates = relevant_phs(embs, cates, int(N)) File "/Users/blabla/PycharmProjects/taxogenpy3/code/local_embedding_training.py", line 62, in relevant_phs sim = utils.cossim(embs[cate_ph], embs[ph]) KeyError: 'determining'

[SasCezar]

I changed the filter_thre parameter in the params.py file to 0.25 and was able to complete the run. However, the results are not nearly as good as in the paper

[DecideToLeave]

Have you solved it? I have a similar problem.

[kekexii]

Have you solved it? I have a similar problem.