cikit-learn 0.24.1 was released a while ago.

https://github.com/chartbeat-labs/textacy/blob/cdedd2351bf2a56e8773ec162e08c3188809d486/setup.cfg#L51

The Nix packages is already at 0.24.1 and this breaks the build.

Do you have any plans to allow scikit-learn > 0.24?

Thanks

what's wrong?

<!-- Describe the problem or confusion here. If you've found a minor mistake and know how to fix it, feel free to skip this issue and submit a pull request instead: https://github.com/chartbeat-labs/textacy/pulls -->

textacy.constants has POS_REGEX_PATTERNS:

textacy.constants.POS_REGEX_PATTERNS['en']['NP'] 

'<DET>? <NUM>* (<ADJ> <PUNCT>? <CONJ>?)* (<NOUN>|<PROPN> <PART>?)+'

What is the equivalent pattern with the newer textacy.extract.matches pattern format?

Should it be in the constants file?

relevant page or section

https://github.com/chartbeat-labs/textacy/blob/cdedd2351bf2a56e8773ec162e08c3188809d486/src/textacy/constants.py#L86

I have tried a sample text that I have seen in this repo to extract SVO triples:

text_str = u'Startup companies create jobs and support innovation. Hilary supports entrepreneurship.'
text = nlp(text_str)
text_ext = textacy.extract.subject_verb_object_triples(text)
list(text_ext)

This runs just fine for me and produces the expected result:

[(companies, create, jobs), (Hilary, supports, entrepreneurship)]

However, I then try to do something a bit more complicated:

text_ada = (u'Ada Lovelace was an English mathematician and' 
            ' writer, chiefly known for her work on'
            ' mechanical general-purpose computer, the'
            ' Analytical Engine. She was the first to'
            ' recognise that the machine had applications'
            ' beyond pure calculation, and published the'
            ' first algorithm intended to be carried out' 
            ' by such a machine. As a result, she is'
            ' sometimes regarded as the first to recognise'
            ' the full potential of a computing machine and'
            ' one of the first computer programmers.')

So I run through the same procedure:

ada = nlp(text_ada)
text_ext = textacy.extract.subject_verb_object_triples(ada)
list(text_ext)

but I get an empty list.

Any thoughts?

Just nitpicking: as a native speaker, I think you wanted to write the second sentence, but I'm not completely sure.

• (Old) Sí, ¡se habla español! / Yes, Spanish is spoken!
• (New) Sí, ¡sé hablar español! / Yes, I can speak Spanish!

By the way: awesome project!

Hi textacy team,

First, thanks for the amazing library.

Any plans to support the new spaCy 3.0 release?

Thanks

Hi @bdewilde, The issue is closed, but I am not sure if it resolved the issue or not. Using "text_with_ws" extracts only the acronym itself, without its full wording. Am I missing something? Can you please let me know if there is a workaround to get it right? Thanks in advance!

Hi all,

As @mzeidhassan mentions, this could indeed be related to a bug in spaCy where calling span.text on an empty Span would result in an IndexError.

There's a few ways of dealing with this:

• Ensure you never define empty spans if you can
• Use self.text_with_ws as a workaround instead

I also would like to confirm that everything was working just fine with spaCy 2.3.2

I think I will also have to do some research as to whether this is a better outcome, stay tuned.

Hi @bdewilde,

I updated spaCy to the latest release '2.3.5', but the acronyms' extraction doesn't seem to be working any longer.

Here is the error I am getting:

IndexError: [E201] Span index out of range.
Traceback:
File "py37/env/lib/python3.7/site-packages/streamlit/script_runner.py", line 332, in _run_script
    exec(code, module.__dict__)
File "py37/env/deep_analysis.py", line 314, in <module>
    acronyms = textacy.extract.acronyms_and_definitions(doc)
File "py37/env/lib/python3.7/site-packages/textacy/extract.py", line 599, in acronyms_and_definitions
    window_ = window.text
File "span.pyx", line 503, in spacy.tokens.span.Span.text.__get__
File "span.pyx", line 190, in spacy.tokens.span.Span.__getitem__

Environment

• operating system: Ubuntu 16.04
• python version: 3.7.9
• spacy version: 2.3.5
• installed spacy models: en_core_web_sm
• textacy version: 0.10.1

Any idea?

Thanks in advance!

So, this is what I was wondering about — do these changes make sense, in practice? My grammar is a bit rusty, but I believe "for textacy's sake" is an adverb clause and not an object of the verb "were". And "(Stack) Overflow" is separate from "God" wrt "Thank". I'm not totally sure if this is a bug or feature, expected or unexpected. Could you help me understand this better? 🙏

Hi @8W9aG, thanks for the PR and your patience. What you're doing in the code makes sense, but I think it would be helpful to see tangible, expected outputs by way of a couple unit tests. Could you add a few representative examples to a test, for reference?

There is a small typo in src/textacy/io/utils.py.

Should read inferred rather than inferrred.

Semi-automated pull request generated by https://github.com/timgates42/meticulous/blob/master/docs/NOTE.md

Do I really only one spelling mistake over all those docs?! 😅 Thanks for the fix.

There is a small typo in src/textacy/io/utils.py.

Should read inferred rather than inferrred.

Semi-automated pull request generated by https://github.com/timgates42/meticulous/blob/master/docs/NOTE.md

Support the handling of prepositions when performing Subject-Verb-Object Information Extraction.

Description

When running Subject-Verb-Object analysis the current program cannot handle looking past a preposition before a noun, and therefore cannot extract the Object correctly when there is a preposition in front of it.

Motivation and Context

I was running this on a test string "Barack Obama was born in Hawaii." and failed to get a Subject-Verb-Object triple.

How Has This Been Tested?

I run a modified version of textacy 0.10.1 that isolates the "subject_verb_object" function and has support for running on spacy 2.1.0. I tested this in this modified version and the input "Barack Obama was born in Hawaii." produces the triple ("Barack Obama", "born in", "Hawaii").

Screenshots (if appropriate):

N/A

Types of changes

• [x] Bug fix (non-breaking change which fixes an issue)
• [ ] New feature (non-breaking change which adds functionality)
• [ ] Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

• [x] My code follows the code style of this project.
• [ ] My change requires a change to the documentation, and I have updated it accordingly.

steps to reproduce

1. initialize doc with text containing the word "didn't"
2. doc._.to_bag_of_terms( ngrams=(1, 2, 3), weighting="count", normalize=None, as_strings=True, filter_stops = False, filter_punct = True, filter_nums = False )

expected vs. actual behavior

Actual behavior: to_bag_of_terms outputs the word "didn't" as: "didn't", "did", "n't". Expected behavior: to_bag_of_terms outputs the word "didn't" as: "didn't"

context

Trying to count n_grams in a block of text. If the block of text contains the word "didn't" and "did", it double counts "did". This also messes up counting n_grams because it counts "n't do" as a bigram instead of "didn't do".

environment

• operating system: linux
• python version: 3.7.9
• spacy version: 2.3.4
• installed spacy models: en_core_web_sm-2.3.1
• textacy version: 0.10.1

Thanks :-)

Hi @nealonhager , I think there are a couple things going on here.

First, ngrams=(1, 2, 3) tells the function to include the combination of unigrams, bigrams, and trigrams in the bag of terms (docs), which by definition will result in overlapping representations of adjacent terms. The terms "did" and "n't" are the unigrams and "didn't" is the bigram that result from the original "didn't".

Second, the word "didn't" (and like contractions) are typically tokenized into "did" and "n't" since they, in a sense, two words mushed together. textacy relies on spacy's tokenization:

>>> import spacy
>>> nlp = spacy.load("en_core_web_sm")
>>> doc = nlp("didn't")
>>> [tok for tok in doc]
[did, n't]

This is all standard and expected behavior. If it doesn't work for your use case, there is functionality for merging tokens: check out merge_spans() here. Using the simple example above:

>>> textacy.spacier.utils.merge_spans([doc[0:2]], doc)
>>> [tok for tok in doc]
[didn't]

steps to reproduce

1. initialize doc with text containing the word "didn't"

2.      ngrams=(1, 2, 3), 
        weighting="count", 
        normalize=None,
        as_strings=True,
        filter_stops = False,
        filter_punct = True,
        filter_nums = False
    )```
   
   

expected vs. actual behavior

Actual behavior: to_bag_of_terms outputs the word "didn't" as: "didn't", "did", "n't". Expected behavior: to_bag_of_terms outputs the word "didn't" as: "didn't"

context

Trying to count n_grams in a block of text. If the block of text contains the word "didn't" and "did", it double counts "did". This also messes up counting n_grams because it counts "n't do" as a bigram instead of "didn't do".

environment

• operating system: linux
• python version: 3.7.9
• spacy version: 2.3.4
• installed spacy models: en_core_web_sm-2.3.1
• textacy version: 0.10.1

Thanks :-)