Spaces:
Sleeping
Sleeping
File size: 26,697 Bytes
da7dbd0 56cf7e3 26e3944 da7dbd0 26e3944 da7dbd0 1ce1659 badcb49 da7dbd0 d952fbe da7dbd0 a6b0abd d952fbe 26e3944 da7dbd0 1ce1659 da7dbd0 1ce1659 da7dbd0 26e3944 a6b0abd 56cf7e3 a6b0abd 56cf7e3 d952fbe a6b0abd 56cf7e3 a6b0abd da7dbd0 a6b0abd 26e3944 a6b0abd 26e3944 a6b0abd 1ce1659 da7dbd0 a6b0abd da7dbd0 a6b0abd da7dbd0 1ce1659 da7dbd0 d952fbe 26e3944 56cf7e3 26e3944 d952fbe 56cf7e3 d952fbe da7dbd0 d952fbe 26e3944 d952fbe 26e3944 d952fbe 26e3944 d952fbe 26e3944 d952fbe 26e3944 d952fbe da7dbd0 d952fbe da7dbd0 d952fbe b489aea d952fbe da7dbd0 d952fbe da7dbd0 d952fbe 1ce1659 26e3944 56cf7e3 26e3944 56cf7e3 26e3944 56cf7e3 26e3944 56cf7e3 26e3944 d952fbe 26e3944 d952fbe 56cf7e3 26e3944 b489aea a6b0abd d952fbe 26e3944 d952fbe 1ce1659 26e3944 56cf7e3 26e3944 56cf7e3 da7dbd0 56cf7e3 |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 |
from difflib import SequenceMatcher
from src.application.image.image_detection import detect_image_by_ai_model, detect_image_by_reverse_search, detect_image_from_news_image
from src.application.text.entity import apply_highlight, highlight_entities
from src.application.text.helper import extract_equal_text
from src.application.text.model_detection import detect_text_by_ai_model
from src.application.text.preprocessing import split_into_paragraphs
from src.application.text.search_detection import check_human, detect_text_by_relative_search
class NewsVerification():
def __init__(self):
self.news_text = ""
self.news_title = ""
self.news_content = ""
self.news_image = ""
self.text_prediction_label:list[str] = []
self.text_prediction_score:list[float] = []
self.text_referent_url:list[str] = []
self.image_prediction_label:list[str] = []
self.image_prediction_score:list[str] = []
self.image_referent_url:list[str] = []
self.news_prediction_label = ""
self.news_prediction_score = -1
self.found_img_url:list[str] = ["https://ichef.bbci.co.uk/ace/standard/819/cpsprodpb/8acc/live/86282470-defb-11ef-ba00-65100a906e68.jpg"]
self.aligned_sentences:list[dict] = []
self.is_paraphrased:list[bool] = []
self.ordinary_user_table:list = []
self.fact_checker_table:list = []
self.governor_table:list = []
def load_news(self, news_title, news_content, news_image):
self.news_text = news_title + "\n\n" + news_content
self.news_title = news_title
self.news_content = news_content
self.news_image = news_image
def determine_text_origin(self):
"""
Determines the origin of the given text based on paraphrasing detection and human authorship analysis.
Args:
text: The input text to be analyzed.
Returns:
str: The predicted origin of the text:
- "HUMAN": If the text is likely written by a human.
- "MACHINE": If the text is likely generated by a machine.
"""
print("CHECK TEXT:")
print("\tFrom search engine:")
# Classify by search engine
input_sentences = split_into_paragraphs(self.news_text)
current_index = 0
previous_paraphrase = None
ai_sentence = {
"input_sentence": "",
"matched_sentence": "",
"label": "",
"similarity": None,
"paraphrase": False,
"url": "",
}
for index, sentence in enumerate(input_sentences):
print(f"-------index = {index}-------")
print(f"current_sentence = {input_sentences[index]}")
if current_index >= len(input_sentences):
break
if current_index >= index and index != 0 and index != len(input_sentences) - 1:
continue
paraphrase, text_url, searched_sentences, img_urls, current_index = detect_text_by_relative_search(input_sentences, index)
if paraphrase is False:
# add sentence to ai_sentence
if ai_sentence["input_sentence"] != "":
ai_sentence["input_sentence"] += "<br>"
ai_sentence["input_sentence"] += sentence
if index == len(input_sentences) - 1:
# add ai_sentences to align_sentences
text_prediction_label, text_prediction_score = detect_text_by_ai_model(ai_sentence["input_sentence"])
ai_sentence["label"] = text_prediction_label
ai_sentence["similarity"] = text_prediction_score
self.aligned_sentences.append(ai_sentence)
else:
if previous_paraphrase is False or previous_paraphrase is None:
# add ai_sentences to align_sentences
if ai_sentence["input_sentence"] != "" or current_index >= len(input_sentences):
text_prediction_label, text_prediction_score = detect_text_by_ai_model(ai_sentence["input_sentence"])
ai_sentence["label"] = text_prediction_label
ai_sentence["similarity"] = text_prediction_score
self.aligned_sentences.append(ai_sentence)
# reset
ai_sentence = {
"input_sentence": "",
"matched_sentence": "",
"label": "",
"similarity": None,
"paraphrase": False,
"url": "",
}
# add searched_sentences to align_sentences
if searched_sentences["input_sentence"] != "":
self.found_img_url.extend(img_urls)
if check_human(searched_sentences):
searched_sentences["label"] = "HUMAN"
else:
searched_sentences["label"] = "MACHINE"
self.aligned_sentences.append(searched_sentences)
previous_paraphrase = paraphrase
def detect_image_origin(self):
print("CHECK IMAGE:")
if self.news_image is None:
self.image_prediction_label = "UNKNOWN"
self.image_prediction_score = 0.0
self.image_referent_url = None
return
for image in self.found_img_url:
print(f"\tfound_img_url: {image}")
matched_url, similarity = detect_image_from_news_image(self.news_image, self.found_img_url)
if matched_url is not None:
print(f"matching image: {matched_url}\nsimilarity: {similarity}\n")
self.image_prediction_label = "HUMAN"
self.image_prediction_score = similarity
self.image_referent_url = matched_url
return
matched_url, similarity = detect_image_by_reverse_search(self.news_image)
if matched_url is not None:
print(f"matching image: {matched_url}\nsimilarity: {similarity}\n")
self.image_prediction_label = "HUMAN"
self.image_prediction_score = similarity
self.image_referent_url = matched_url
return
detected_label, score = detect_image_by_ai_model(self.news_image)
if detected_label:
print(f"detected_label: {detected_label} ({score})")
self.image_prediction_label = detected_label
self.image_prediction_score = score
self.image_referent_url = None
return
self.image_prediction_label = "UNKNOWN"
self.image_prediction_score = 50
self.image_referent_url = None
def determine_news_origin(self):
if self.text_prediction_label == "MACHINE":
text_prediction_score = 100 - self.text_prediction_score
elif self.text_prediction_label == "UNKNOWN":
text_prediction_score = 50
else:
text_prediction_score = self.text_prediction_score
if self.image_prediction_label == "MACHINE":
image_prediction_score = 100 - self.image_prediction_score
elif self.image_prediction_label == "UNKNOWN":
image_prediction_score = 50
else:
image_prediction_score = self.image_prediction_score
news_prediction_score = (text_prediction_score + image_prediction_score) / 2
if news_prediction_score > 50:
self.news_prediction_score = news_prediction_score
self.news_prediction_label = "HUMAN"
else:
self.news_prediction_score = 100 - news_prediction_score
self.news_prediction_label = "MACHINE"
def generate_analysis_report(self):
self.determine_text_origin()
self.detect_image_origin()
def analyze_details(self):
ordinary_user_table = self.create_ordinary_user_table()
fact_checker_table = self.create_fact_checker_table()
governor_table = self.create_governor_table()
return ordinary_user_table, fact_checker_table, governor_table
def get_text_urls(self):
return set(self.text_referent_url)
def compare_sentences(self, sentence_1, sentence_2, position, color):
"""
Compares two sentences and identifies common phrases, outputting their start and end positions.
Args:
sentence_1: The first sentence (string).
sentence_2: The second sentence (string).
Returns:
A list of dictionaries, where each dictionary represents a common phrase and contains:
- "phrase": The common phrase (string).
- "start_1": The starting index of the phrase in sentence_1 (int).
- "end_1": The ending index of the phrase in sentence_1 (int).
- "start_2": The starting index of the phrase in sentence_2 (int).
- "end_2": The ending index of the phrase in sentence_2 (int).
Returns an empty list if no common phrases are found. Handles edge cases like empty strings.
"""
if not sentence_1 or not sentence_2: # Handle empty strings
return []
s = SequenceMatcher(None, sentence_1, sentence_2)
common_phrases = []
for block in s.get_matching_blocks():
if block.size > 0: # Ignore zero-length matches
start_1 = block.a
end_1 = block.a + block.size
start_2 = block.b
end_2 = block.b + block.size
phrase = sentence_1[start_1:end_1] # Or sentence_2[start_2:end_2], they are the same
common_phrases.append({
"phrase": phrase,
"start_1": start_1 + position,
"end_1": end_1 + position,
"start_2": start_2,
"end_2": end_2,
"color": color,
})
position += len(sentence_1)
return common_phrases, position
def create_fact_checker_table(self):
rows = []
max_length = 30 # TODO: put this in configuration
rows.append(self.format_image_fact_checker_row(max_length))
for aligned_sentence in self.aligned_sentences:
if "input_sentence" not in aligned_sentence:
continue
# Get index of equal phrases in input and source sentences
equal_idx_1, equal_idx_2 = extract_equal_text(
aligned_sentence["input_sentence"],
aligned_sentence["matched_sentence"],
)
# Get entity-words (in pair) with colors
entities_with_colors = highlight_entities(
aligned_sentence["input_sentence"],
aligned_sentence["matched_sentence"],
)
self.fact_checker_table.append(
[
aligned_sentence,
equal_idx_1,
equal_idx_2,
entities_with_colors,
]
)
for row in self.fact_checker_table:
formatted_row = self.format_text_fact_checker_row(row, max_length)
rows.append(formatted_row)
table = "\n".join(rows)
return f"""
<h5>Comparison between input news and source news</h5>
<table border="1" style="width:100%; text-align:left; border-collapse:collapse;">
<thead>
<tr>
<th>Input news</th>
<th>Source (URL provided in Originality column correspondingly)</th>
<th>Forensic</th>
<th>Originality</th>
</tr>
</thead>
<tbody>
{table}
</tbody>
</table>
<style>
"""
def format_text_fact_checker_row(self, row, max_length=30):
if row[0]["input_sentence"] == "":
return ""
if row[0]["matched_sentence"] != "": # source is not empty
# highlight entities
input_sentence, highlight_idx_input = apply_highlight(row[0]["input_sentence"], row[3], "input")
source_sentence, highlight_idx_source = apply_highlight(row[0]["matched_sentence"], row[3], "source")
# Color overlapping words
input_sentence = self.color_text(input_sentence, row[1], highlight_idx_input) # text, index of highlight words
source_sentence = self.color_text(source_sentence, row[2], highlight_idx_source) # text, index of highlight words
input_sentence = input_sentence.replace("span_style", "span style").replace("1px_4px", "1px 4px")
source_sentence = source_sentence.replace("span_style", "span style").replace("1px_4px", "1px 4px")
else:
input_sentence = row[0]["input_sentence"]
source_sentence = row[0]["matched_sentence"]
label = row[0]["label"]
score = row[0]["similarity"]
url = row[0]["url"] #
short_url = self.shorten_url(url, max_length)
source_text_url = f"""<a href="{url}">{short_url}</a>"""
return f"""
<tr>
<td>{input_sentence}</td>
<td>{source_sentence}</td>
<td>{label}<br>({score*100:.2f}%)</td>
<td>{source_text_url}</td>
</tr>
"""
def format_image_fact_checker_row(self, max_length=30):
if self.image_referent_url is not None or self.image_referent_url != "":
source_image = f"""<img src="{self.image_referent_url}" width="200" height="150">"""
short_url = self.shorten_url(self.image_referent_url, max_length)
source_image_url = f"""<a href="{self.image_referent_url}">{short_url}</a>"""
else:
source_image = "Image not found"
source_image_url = ""
return f"""<tr><td>input image</td><td>{source_image}</td><td>{self.image_prediction_label}<br>({self.image_prediction_score:.2f}%)</td><td>{source_image_url}</td></tr>"""
def create_ordinary_user_table(self):
rows = []
max_length = 30 # TODO: put this in configuration
rows.append(self.format_image_ordinary_user_row(max_length))
rows.append(self.format_text_ordinary_user_row(max_length))
table = "\n".join(rows)
return f"""
<h5>Comparison between input news and source news</h5>
<table border="1" style="width:100%; text-align:left; border-collapse:collapse;">
<thead>
<tr>
<th>Input news</th>
<th>Forensic</th>
<th>Originality</th>
</tr>
</thead>
<tbody>
{table}
</tbody>
</table>
<style>
"""
def format_text_ordinary_user_row(self, max_length=30):
input_sentences = ""
source_text_urls = ""
label = ""
scores = 0
sentence_count = 0
for index, row in enumerate(self.aligned_sentences):
if row["input_sentence"] == "":
continue
input_sentences += row["input_sentence"]
label = self.aligned_sentences[index]["label"]
if label == "HUMAN":
score = self.aligned_sentences[index]["similarity"]
if label == "MACHINE":
score = 1 - self.aligned_sentences[index]["similarity"]
scores += score
url = self.aligned_sentences[index]["url"] #
short_url = self.shorten_url(url, max_length)
source_text_urls += f"""<a href="{url}">{short_url}</a><br>"""
sentence_count += 1
if scores == 0:
label = "UNKNOWN"
else:
scores /= sentence_count
if scores > 0.5:
label = "HUMAN"
else:
label = "MACHINE"
scores = 1 - scores
return f"""
<tr>
<td>{input_sentences}</td>
<td>{label}<br>({scores*100:.2f}%)</td>
<td>{source_text_urls}</td>
</tr>
"""
def format_image_ordinary_user_row(self, max_length=30):
if self.image_referent_url is not None or self.image_referent_url != "":
source_image = f"""<img src="{self.image_referent_url}" width="200" height="150">"""
short_url = self.shorten_url(self.image_referent_url, max_length)
source_image_url = f"""<a href="{self.image_referent_url}">{short_url}</a>"""
else:
source_image = "Image not found"
source_image_url = ""
return f"""<tr><td>input image</td><td>{source_image}</td><td>{self.image_prediction_label}<br>({self.image_prediction_score:.2f}%)</td><td>{source_image_url}</td></tr>"""
def create_governor_table(self):
rows = []
max_length = 30 # TODO: put this in configuration
rows.append(self.format_image_governor_row(max_length))
for aligned_sentence in self.aligned_sentences:
if "input_sentence" not in aligned_sentence:
continue
# Get index of equal phrases in input and source sentences
equal_idx_1, equal_idx_2 = extract_equal_text(
aligned_sentence["input_sentence"],
aligned_sentence["matched_sentence"],
)
# Get entity-words (in pair) with colors
entities_with_colors = highlight_entities(
aligned_sentence["input_sentence"],
aligned_sentence["matched_sentence"],
)
self.governor_table.append(
[
aligned_sentence,
equal_idx_1,
equal_idx_2,
entities_with_colors,
]
)
formatted_row = self.format_text_governor_row(max_length)
rows.append(formatted_row)
table = "\n".join(rows)
return f"""
<h5>Comparison between input news and source news</h5>
<table border="1" style="width:100%; text-align:left; border-collapse:collapse;">
<col style="width: 150px;"> <col style="width: 150px;"> <col style="width: 50px;"> <col style="width: 75px;">
<thead>
<tr>
<th>Input news</th>
<th>Source (URL provided in Originality column correspondingly)</th>
<th>Forensic</th>
<th>Originality</th>
</tr>
</thead>
<tbody>
{table}
</tbody>
</table>
<style>
"""
def format_text_governor_row(self, max_length=30):
input_sentences = ""
source_sentences = ""
source_text_urls = ""
label = ""
scores = 0
sentence_count = 0
entity_count = 0
for row in self.governor_table:
print(f"governor_row: {row}")
if row[0]["input_sentence"] == "":
continue
if row[0]["matched_sentence"] != "": # source is not empty
# highlight entities
input_sentence, highlight_idx_input = apply_highlight(row[0]["input_sentence"], row[3], "input", entity_count)
source_sentence, highlight_idx_source = apply_highlight(row[0]["matched_sentence"], row[3], "source", entity_count)
entity_count += len(row[3])
# Color overlapping words
input_sentence = self.color_text(input_sentence, row[1], highlight_idx_input) # text, index of highlight words
source_sentence = self.color_text(source_sentence, row[2], highlight_idx_source) # text, index of highlight words
input_sentence = input_sentence.replace("span_style", "span style").replace("1px_4px", "1px 4px")
source_sentence = source_sentence.replace("span_style", "span style").replace("1px_4px", "1px 4px")
else:
input_sentence = row[0]["input_sentence"]
source_sentence = row[0]["matched_sentence"]
input_sentences += input_sentence
source_sentences += source_sentence
score = row[0]["similarity"]
label = row[0]["label"]
if label == "HUMAN":
score = row[0]["similarity"]
if label == "MACHINE":
score = 1 - row[0]["similarity"]
scores += score
url = row[0]["url"]
short_url = self.shorten_url(url, max_length)
source_text_urls += f"""<a href="{url}">{short_url}</a><br>"""
sentence_count += 1
if scores == 0:
label = "UNKNOWN"
else:
scores /= sentence_count
if scores > 0.5:
label = "HUMAN"
else:
label = "MACHINE"
scores = 1 - scores
return f"""
<tr>
<td>{input_sentences}</td>
<td>{source_sentences}</td>
<td>{label}<br>({score*100:.2f}%)</td>
<td>{source_text_urls}</td>
</tr>
"""
def format_image_governor_row(self, max_length=30):
if self.image_referent_url is not None or self.image_referent_url != "":
source_image = f"""<img src="{self.image_referent_url}" width="200" height="150">"""
short_url = self.shorten_url(self.image_referent_url, max_length)
source_image_url = f"""<a href="{self.image_referent_url}">{short_url}</a>"""
else:
source_image = "Image not found"
source_image_url = ""
return f"""<tr><td>input image</td><td>{source_image}</td><td>{self.image_prediction_label}<br>({self.image_prediction_score:.2f}%)</td><td>{source_image_url}</td></tr>"""
def shorten_url(self, url, max_length=30):
if url is None:
return ""
if len(url) > max_length:
short_url = url[:max_length] + "..."
else:
short_url = url
return short_url
def color_text(self, text, colored_idx, highlighted_idx):
paragraph = ""
words = text.split()
starts, ends = self.extract_starts_ends(colored_idx)
starts, ends = self.filter_indices(starts, ends, highlighted_idx)
previous_end = 0
for start, end in zip(starts, ends):
paragraph += " ".join(words[previous_end:start])
equal_words = " ".join(words[start:end])
paragraph += f" <span style='color:#00FF00;'>{equal_words}</span> "
previous_end = end
# Some left words due to the punctuation separated from
# the highlighting text
equal_words = " ".join(words[previous_end:])
print(f"starts_2: {previous_end}")
print(f"ends_2: {len(words)-1}")
print(f"equal_words: {words[previous_end:]}")
paragraph += f" <span style='color:#00FF00;'>{equal_words}</span> "
return paragraph
def extract_starts_ends(self, colored_idx):
starts = []
ends = []
for index in colored_idx:
starts.append(index['start'])
ends.append(index['end'])
return starts, ends
def filter_indices(self, starts, ends, ignore_indices):
"""
Filters start and end indices to exclude any indices present in the ignore_indices list.
Args:
starts: A list of starting indices.
ends: A list of ending indices. Must be the same length as starts.
ignore_indices: A list of indices to exclude.
Returns:
A tuple containing two new lists: filtered_starts and filtered_ends.
Returns empty lists if the input is invalid or if all ranges are filtered out.
Prints error messages for invalid input.
Examples:
starts = [0, 5, 10]
ends = [3, 7, 12]
ignore_indices = [1, 2, 11, 17]
# Output:
starts = [0, 3, 5, 10, 12]
ends = [0, 3, 7, 10, 12]
"""
if len(starts) != len(ends):
print("Error: The 'starts' and 'ends' lists must have the same length.")
return [], []
filtered_starts = []
filtered_ends = []
for i in range(len(starts)):
start = starts[i]
end = ends[i]
if end < start:
print(f"Error: End index {end} is less than start index {start} at position {i}.")
return [], []
start_end = list(range(start, end + 1, 1))
start_end = list(set(start_end) - set(ignore_indices))
new_start, new_end = self.extract_sequences(start_end)
filtered_starts.extend(new_start)
filtered_ends.extend(new_end)
return filtered_starts, filtered_ends
def extract_sequences(self, numbers):
if len(numbers) == 1:
return [numbers[0]], [numbers[0]]
numbers.sort()
starts = []
ends = []
for i, number in enumerate(numbers):
if i == 0:
start = number
end = number
continue
if number - 1 == numbers[i-1]:
end = number
else:
starts.append(start)
ends.append(end + 1)
start = number
end = number
if i == len(numbers) - 1:
starts.append(start)
ends.append(end + 1)
return starts, ends |