File size: 6,934 Bytes
9ff961b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
import spacy
import gradio as gr
import json
from typing import Dict, List, Tuple, Any

from zshot import PipelineConfig
from zshot.linker import LinkerSMXM
from zshot.utils.data_models import Entity

from spacy.cli import download
download("en_core_web_sm")

# Function to load the NER model
def load_model(entity_data):
    entities = [
        Entity(
            name=entity["name"],
            description=entity["description"],
            vocabulary=entity.get("vocabulary")
        ) for entity in entity_data
    ]
    
    nlp = spacy.blank("en")
    nlp_config = PipelineConfig(
        linker=LinkerSMXM(model_name="disi-unibo-nlp/openbioner-base"),
        entities=entities,
        device='cpu'  # Change to 'cpu' if GPU not available
    )
    nlp.add_pipe("zshot", config=nlp_config, last=True)
    
    return nlp

# Default entities - focusing on BACTERIUM example
default_entities = [
    {
        "name": "BACTERIUM",
        "description": "A bacterium refers to a type of microorganism that can exist as a single cell and may cause infections or play a role in various biological processes. Examples include species like Streptococcus pneumoniae and Streptomyces ahygroscopicus.",
    }
]

# Initialize model with default entities
nlp = load_model(default_entities)

# Function to create HTML visualization of entities
def get_entity_html(doc) -> str:
    colors = {
        "BACTERIUM": "#8dd3c7",
        "CHEMICAL": "#fb8072", 
        "DISEASE": "#80b1d3",
        "GENE": "#fdb462",
        "SPECIES": "#b3de69"
    }
    
    html_parts = []
    last_idx = 0
    
    # Display text with highlighted entities
    for ent in doc.ents:
        # Add text before the entity
        html_parts.append(doc.text[last_idx:ent.start_char])
        
        # Add the highlighted entity
        color = colors.get(ent.label_, "#ddd")
        html_parts.append(
            f'<span style="background-color: {color}; padding: 0.2em 0.3em; '
            f'border-radius: 0.35em; margin: 0 0.1em; font-weight: bold; color: #000;">'
            f'{doc.text[ent.start_char:ent.end_char]}'
            f'<span style="font-size: 0.8em; font-weight: bold; margin-left: 0.5em">{ent.label_}</span>'
            f'</span>'
        )
        
        # Update the last index
        last_idx = ent.end_char
    
    # Add any remaining text
    html_parts.append(doc.text[last_idx:])
    
    # Wrap the result in a div with dark theme styling
    return f'<div style="line-height: 1.5; padding: 10px; background: #222; color: #fff; border-radius: 5px;">{"".join(html_parts)}</div>'

# Function to get entity details including spans
def get_entity_details(doc) -> List[Dict[str, Any]]:
    entity_details = []
    for ent in doc.ents:
        entity_details.append({
            "text": ent.text,
            "type": ent.label_,
            "start": ent.start_char,
            "end": ent.end_char
        })
    return entity_details

# Main processing function
def process_text(text: str, entities_json: str) -> Tuple[str, List[Dict[str, Any]]]:
    global nlp
    
    # Update model if entities have changed
    try:
        entities = json.loads(entities_json)
        nlp = load_model(entities)
    except json.JSONDecodeError:
        return "Error: Invalid JSON in entity configuration", []
    
    # Process the text with the NER model
    doc = nlp(text)
    
    # Generate visualization HTML
    html_output = get_entity_html(doc)
    
    # Get detailed entity information including spans
    entity_details = get_entity_details(doc)
    
    return html_output, entity_details

# Set theme to dark
theme = gr.themes.Soft(
    primary_hue="blue",
    secondary_hue="slate",
    neutral_hue="slate",
    text_size=gr.themes.sizes.text_md,
).set(
    body_background_fill="#1a1a1a",
    background_fill_primary="#222",
    background_fill_secondary="#333",
    border_color_primary="#444",
    block_background_fill="#222",
    block_label_background_fill="#333",
    block_label_text_color="#fff",
    block_title_text_color="#fff",
    body_text_color="#fff",
    button_primary_background_fill="#2563eb",
    button_primary_text_color="#fff",
    input_background_fill="#333",
    input_border_color="#555",
    input_placeholder_color="#888",
    panel_background_fill="#222",
    slider_color="#2563eb",
)

# Create Gradio interface with dark theme
with gr.Blocks(title="Named Entity Recognition", theme=theme) as demo:
    gr.Markdown("# OpenBioNER - Demo")
    
    # First row: Entity Definitions
    with gr.Row():
        entities_input = gr.Code(
            label="Entity Definitions (JSON)",
            language="json", 
            value=json.dumps(default_entities, indent=2),
            lines=6
        )
    
    # Second row: Input text and examples side by side
    with gr.Row():
        # Left side - Input text
        with gr.Column():
            text_input = gr.Textbox(
                label="Text to analyze",
                placeholder="Enter text to analyze...",
                value="Impact of cofactor - binding loop mutations on thermotolerance and activity of E. coli transketolase",
                lines=3
            )
            
            analyze_btn = gr.Button("Analyze Text", variant="primary")
        
        # Right side - Example texts
        with gr.Column():
            gr.Markdown("### Quick Examples")
            example1_btn = gr.Button("E. coli research")
            example2_btn = gr.Button("Bacterial infection case")
            example3_btn = gr.Button("Multiple bacterial species")
    
    # Third row: Output visualization and spans side by side
    with gr.Row():
        # Left side - Highlighted text output
        with gr.Column():
            gr.Markdown("### Recognized Entities")
            result_html = gr.HTML()
        
        # Right side - Entity spans details
        with gr.Column():
            gr.Markdown("### Entity Details with Spans")
            entity_details = gr.JSON()
    
    # Set up event handlers for the analyze button
    analyze_btn.click(
        fn=process_text,
        inputs=[text_input, entities_input],
        outputs=[result_html, entity_details]
    )
    
    # Set up event handlers for example buttons
    example1_btn.click(
        fn=lambda: "Impact of cofactor - binding loop mutations on thermotolerance and activity of E. coli transketolase",
        inputs=None,
        outputs=text_input
    )
    
    example2_btn.click(
        fn=lambda: "The patient was diagnosed with pneumonia caused by Streptococcus pneumoniae and treated with antibiotics for 7 days.",
        inputs=None,
        outputs=text_input
    )
    
    example3_btn.click(
        fn=lambda: "We compared growth rates of E. coli, B. subtilis and S. aureus in various media containing different carbon sources.",
        inputs=None,
        outputs=text_input
    )

if __name__ == "__main__":
    demo.launch()