File size: 7,265 Bytes
9c6961c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import gradio as gr
import json
import random
import os
from datetime import datetime

# --- PATH CONFIGURATION ---
DATA_PATH = "/home/mshahidul/readctrl/data/data_annotator_data/crowdsourcing_input_en_v2.json"
SAVE_ROOT = "/home/mshahidul/readctrl/data/annotators_validate_data"
QUESTIONS_FILE = "/home/mshahidul/readctrl/code/interface/sp50_questions.json"

# --- SESSION CONFIGURATION ---
NUM_QUESTIONS = 30        
NUM_DUPLICATES = 4       
NUM_LITERACY_QUERIES = 10 
DUPLICATE_INTERVAL = 8   

# --- ANNOTATION GUIDE TEXT ---
GUIDE_HTML = """
<div style="background-color: #f9f9f9; padding: 15px; border-left: 6px solid #2196F3; border-radius: 4px;">
    <h3>Rating Guide: Medical Text Difficulty</h3>
    <p>Please rate the difficulty of the documents based on the following scale:</p>
    <table style="width:100%; border-collapse: collapse; text-align: left;">
        <tr style="background-color: #e3f2fd;">
            <th style="padding: 8px; border: 1px solid #ddd;">Score</th>
            <th style="padding: 8px; border: 1px solid #ddd;">Description</th>
        </tr>
        <tr>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>1 - 2</b></td>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>Very Easy:</b> Clear language, no medical jargon. Like a 5th-grade textbook.</td>
        </tr>
        <tr>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>3 - 4</b></td>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>Easy:</b> Common medical terms (e.g., "fever", "heart") used in simple sentences.</td>
        </tr>
        <tr>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>5 - 6</b></td>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>Moderate:</b> Some technical terms. Requires focused reading but understandable.</td>
        </tr>
        <tr>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>7 - 8</b></td>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>Hard:</b> Heavy use of medical jargon. Read like a clinical report.</td>
        </tr>
        <tr>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>9 - 10</b></td>
            <td style="padding: 8px; border: 1px solid #ddd;"><b>Very Hard:</b> Specialist-level text. Extremely dense and difficult to follow.</td>
        </tr>
    </table>
</div>
"""

def load_questions():
    with open(QUESTIONS_FILE, "r") as f:
        all_q = json.load(f)
    return random.sample(all_q, min(NUM_LITERACY_QUERIES, len(all_q)))

class AnnotationSession:
    def __init__(self, dataset, questions):
        base_samples = random.sample(dataset, NUM_QUESTIONS)
        self.queue = list(base_samples)
        for i in range(NUM_DUPLICATES):
            self.queue.insert(DUPLICATE_INTERVAL + i, base_samples[i])
        
        self.current_index = 0
        self.results = []
        self.questions = questions
        self.session_folder = None 

with open(DATA_PATH, "r") as f:
    full_dataset = json.load(f)

session = AnnotationSession(full_dataset, load_questions())

def start_and_save_literacy(*answers):
    timestamp = datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
    session_folder = os.path.join(SAVE_ROOT, timestamp)
    os.makedirs(session_folder, exist_ok=True)
    session.session_folder = session_folder 

    literacy_data = []
    for i, ans in enumerate(answers):
        q_info = session.questions[i]
        literacy_data.append({
            "question_id": q_info['id'],
            "question_text": q_info['question'],
            "user_answer": ans,
            "is_correct": ans == q_info['correct']
        })
    
    with open(os.path.join(session_folder, "literacy_results.json"), "w") as f:
        json.dump(literacy_data, f, indent=4)
    
    first_pair = session.queue[0]
    return (
        gr.update(visible=False), 
        gr.update(visible=True),  
        first_pair['original_doc'], 
        first_pair['wiki_anchor'],
        f"Item 1 of {len(session.queue)}"
    )

def submit_rating(doc_slider, wiki_slider):
    # 1. Capture the current result
    current_pair = session.queue[session.current_index]
    session.results.append({
        "original_index": current_pair.get('index', 'unknown'),
        "queue_position": session.current_index,
        "doc_rating": doc_slider,
        "wiki_rating": wiki_slider,
        "timestamp": datetime.now().strftime("%Y-%m-%d %H:%M:%S")
    })
    
    # 2. Incremental Save: Write to file immediately on every click
    annotation_file = os.path.join(session.session_folder, "annotation_results.json")
    with open(annotation_file, "w") as f:
        json.dump(session.results, f, indent=4)
    
    # 3. Show Pop-up Notification (Gradio Info Toast)
    gr.Info(f"Progress Saved: Item {session.current_index + 1} recorded.")

    # Increment index
    session.current_index += 1
    
    # 4. Check if session is finished
    if session.current_index < len(session.queue):
        next_pair = session.queue[session.current_index]
        return (
            next_pair['original_doc'], 
            next_pair['wiki_anchor'], 
            f"Item {session.current_index + 1} of {len(session.queue)}",
            5, 5  # Reset sliders to neutral middle value
        )
    else:
        # Final update for the UI when done
        return (
            "✅ ALL TASKS COMPLETED", 
            "The data has been saved to your session folder. You may close this tab.", 
            "Status: Finished", 
            0, 0
        )

# --- UI ---
with gr.Blocks(theme=gr.themes.Soft()) as demo:
    gr.Markdown("# Medical Text Readability Annotation")
    
    # Instructions available at all times via Accordion
    with gr.Accordion("See Annotation Instructions & Scale Guide", open=True):
        gr.HTML(GUIDE_HTML)

    with gr.Column(visible=True) as intro_box:
        gr.Markdown(f"### Pre-Task: Health Literacy Check ({NUM_LITERACY_QUERIES} Questions)")
        literacy_inputs = []
        for q in session.questions:
            radio = gr.Radio(choices=q['options'], label=q['question'])
            literacy_inputs.append(radio)
        btn_start = gr.Button("Start Annotation", variant="primary")

    with gr.Column(visible=False) as task_box:
        progress = gr.Label(label="Progress")
        with gr.Row():
            with gr.Column():
                doc_display = gr.Textbox(interactive=False, lines=12, label="Document D (Medical Text)")
                doc_slider = gr.Slider(1, 10, step=1, label="Difficulty (1: Simple → 10: Technical)", value=0)
            with gr.Column():
                wiki_display = gr.Textbox(interactive=False, lines=12, label="Document W (Wikipedia Text)")
                wiki_slider = gr.Slider(1, 10, step=1, label="Difficulty (1: Simple → 10: Technical)", value=0)
        btn_submit = gr.Button("Submit & Next", variant="primary")

    btn_start.click(
        start_and_save_literacy, 
        inputs=literacy_inputs, 
        outputs=[intro_box, task_box, doc_display, wiki_display, progress]
    )
    
    btn_submit.click(
        submit_rating,
        inputs=[doc_slider, wiki_slider],
        outputs=[doc_display, wiki_display, progress, doc_slider, wiki_slider]
    )

demo.launch(share=True)