Spaces:
Sleeping
Sleeping
File size: 2,242 Bytes
9d8ab14 2c1ec44 9d8ab14 2c1ec44 9d8ab14 2c1ec44 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 |
import gradio as gr
from transformers import pipeline
from collections import defaultdict
ner_pipeline = pipeline("ner", model="syubraj/spanish_bert_based_ner")
ENTITY_COLORS = {
"B-PER": "#FF5733",
"I-PER": "#FF8D1A",
"B-LOC": "#33FF57",
"I-LOC": "#2EB82E",
"B-ORG": "#3380FF",
"I-ORG": "#1A66FF",
"O": "#E0E0E0"
}
def highlight_entities(text):
"""
This function takes a Spanish text input and applies Named Entity Recognition (NER) using a pre-trained model.
The recognized entities are highlighted with distinct colors based on their categories.
A legend is also generated to indicate the meaning of each color.
Parameters:
text (str): The input Spanish text.
Returns:
str: The HTML formatted output with highlighted entities and a legend.
"""
entities = ner_pipeline(text)
formatted_text = ""
last_idx = 0
for ent in entities:
word = ent['word']
start, end = ent['start'], ent['end']
entity_label = ent['entity']
color = ENTITY_COLORS.get(entity_label, "#E0E0E0")
formatted_text += text[last_idx:start] + f'<span style="background-color:{color}; padding:2px; border-radius:5px;">{word}</span>'
last_idx = end
formatted_text += text[last_idx:]
legend_html = "<div><b>Legend:</b><br>"
for label, color in ENTITY_COLORS.items():
legend_html += f'<span style="background-color:{color}; padding:2px 5px; border-radius:5px; margin-right:5px;">{label}</span>'
legend_html += "</div><br>"
return legend_html + formatted_text
example_text = "Elon Musk vive en Estados Unidos y es dueño de SpaceX, Tesla y Starlink."
grn = gr.Interface(
fn=highlight_entities,
inputs=gr.Textbox(label="Enter Spanish Text", value=example_text),
outputs=gr.HTML(label="NER Highlighted Text"),
title="Spanish Named Entity Recognition",
description="This interactive demo performs Named Entity Recognition (NER) on Spanish text. Recognized entities such as persons, locations, and organizations are highlighted in distinct colors for better readability. A legend is provided to help interpret the color coding.",
)
grn.launch()
|