skblaz / rakun Goto Github PK

Hello,
I can't find the definition of function hairball_plot, where could I get it?

E NameError: name 'hairball_plot' is not defined
mrakun/init.py:71: NameError

Can I use a specific BERT model with this code? Say I want to use bert-base-uncased to extract keywords. Will it work here?

1. I have tried the code and it is impressive, however I have some difficulties to show in the output not single words, but 2-grams or 3-grams that represent a keyword for specific text. Any info on how to tune the hyperparameters?

2. Concerning the hyperparameters could you explain some of them, such as:

•                "pair_diff_length":2,
  

•                "bigram_count_threshold":2,
  

•                "num_tokens":[1,2],
  

•      "max_similar" : 3, ## n most similar can show up n times
  

•      "max_occurrence" : 3} ## maximum frequency overall
  

Thanks.

Hello, I majored in computer science at Okayama University in Japan.

When I use this algorithm, RaKUn for my NLP task by using Japanese,
I found for loop in "missing connectives" part is a suspected case of bug.

it is following part,
file: __init__.py"
line: 449 ...

for ind in i1_indexes:
                            if ind + 2 in i2_indexes_map:
                                joint_kw = " ".join([
                                    p1, self.raw_text[ind + 1],
                                    self.raw_text[ind + 2]
                                ])
                                final_keywords.append((joint_kw, kw[1]))
                                joint = True

I think if you don't exit this for loop with break, there is a possibility that multiple similar keywords will be added. Is this normal?
Could you please confirm this?

Hi,

when using fasttext in google colab, I'll get the error
module 'gensim.models.fasttext' has no attribute 'load_facebook_vectors'

Any idea?

https://github.com/SkBlaz/rakun/blob/master/mrakun/__init__.py#L29

logging.basicConfig(format='%(asctime)s - %(message)s',
                    datefmt='%d-%b-%y %H:%M:%S')
logging.getLogger().setLevel(logging.INFO)

Using getLogger() without any arguments makes it default to a root logger, and thus
overwrites its level against the users wishes. This makes integrating the package into bigger applications
very difficult as it robs control of the logs from the developer.

Although I'm not aware of what the good practice for setting up loggers for a package is, replacing those lines with the following significantly reduces log clutter:

logging.basicConfig(format='%(asctime)s - %(message)s', datefmt='%d-%b-%y %H:%M:%S')
logging.getLogger(__name__).setLevel(logging.INFO)

I tried to run the sample code provided, but am getting the above error. Unsure how I could debug it, please do help!