Core Module

The core module provides the main functionality for extracting woseco concept associations from text.

WosecoExtractor Class

Usage Example

from wosecopy import WosecoExtractor
import pandas as pd

# Initialize extractor
extractor = WosecoExtractor(language='en', model_size='lg')

# Process a DataFrame
df = pd.DataFrame({'Story': ['Once upon a time...']})
concepts = extractor.get_woseco(df)

# Process a CSV file
result_df = extractor.process_csv(
    'input.csv',
    text_column='Story',
    output_path='output.csv'
)

Key Methods

get_concept_matches

Match concepts between consecutive sentences using semantic similarity.

Algorithm:

Compare all noun pairs between consecutive sentences
Calculate similarity using spaCy word embeddings
Select pair with maximum similarity
Build chain of linked concepts

Example:

concept_list = [
    ['belief', 'village'],
    ['faith', 'people'],
    ['church', 'community']
]
chain = extractor.get_concept_matches(concept_list)
# Returns: ['belief', 'faith', 'church', 'community']

get_woseco

Main function to extract woseco concepts from a DataFrame.

Process:

Split each story into sentences (by periods)
Extract nouns from each sentence
Match concepts between consecutive sentences
Return list of concept chains

Example:

df = pd.DataFrame({
    'Story': [
        'There was belief. Faith followed.',
        'Music played. Songs filled the air.'
    ]
})
concepts = extractor.get_woseco(df)
# Returns: [['belief', 'faith'], ['music', 'song', 'air']]

extract_nouns

Extract and lemmatize nouns from text.

Example:

text = "The dogs are running in the parks."
nouns = extractor.extract_nouns(text)
# Returns: ['dog', 'park']

process_csv

End-to-end processing of a CSV file.

Example:

df = extractor.process_csv(
    'stories.csv',
    text_column='Story',
    output_path='results.csv',
    similarity_threshold=0.5
)