Spaces:
Runtime error
Runtime error
File size: 93,561 Bytes
2cc7f91 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 | <!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Model Details — Phishing Detection</title>
<link rel="stylesheet" href="/static/style.css">
</head>
<body class="models-page">
<div class="container">
<header>
<div class="header-left">
<div class="logo"><a href="/">Phishing Detection System</a></div>
<div class="tagline">Model Performance Details</div>
</div>
<a href="/" class="back-link">← Back</a>
</header>
<section class="page-title-section">
<h1 class="page-title">Model Details</h1>
<p class="page-description">
Performance metrics, feature importance, and configuration details for all 9 machine learning models
used in the phishing detection pipeline. Models span URL features (125), RAW HTML features (77) + engineered (23),
combined features (225), and character-level CNN approaches.
</p>
</section>
<!-- DETECTION PIPELINE -->
<section class="section">
<div class="section-title">Detection Pipeline</div>
<div class="pipeline">
<div class="pipeline-step"><span class="step-number">1</span>URL Input</div>
<div class="pipeline-step"><span class="step-number">2</span>Feature Extraction</div>
<div class="pipeline-step"><span class="step-number">3</span>3 URL Models</div>
<div class="pipeline-step"><span class="step-number">4</span>HTML Download</div>
<div class="pipeline-step"><span class="step-number">5</span>2 HTML + 2 Combined</div>
<div class="pipeline-step"><span class="step-number">6</span>2 CNN Models</div>
<div class="pipeline-step"><span class="step-number">7</span>9-Model Consensus</div>
</div>
</section>
<!-- URL FEATURES -->
<section class="section">
<div class="section-title collapsible-toggle" onclick="toggleFeatures(this)">
URL Features <span class="feature-count">125 features</span>
<span class="toggle-icon">+</span>
</div>
<div class="collapsible-content">
<div class="section-subtitle">All features extracted from the URL string. Hover over any feature to see its description.</div>
<div class="feature-category-label">Length & Structure</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Total character count of the full URL">url_length</div>
<div class="feature-chip" data-tip="Character count of the domain name only">domain_length</div>
<div class="feature-chip" data-tip="Character count of the URL path component">path_length</div>
<div class="feature-chip" data-tip="Character count of the query string">query_length</div>
<div class="feature-chip" data-tip="URL length bucket: 0=short (<40), 1=medium, 2=long, 3=very long (>120)">url_length_category</div>
<div class="feature-chip" data-tip="Domain length bucket: 0=short (<10), 1=medium, 2=long, 3=very long (>30)">domain_length_category</div>
</div>
<div class="feature-category-label">Character Counts</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Number of dots (.) in the full URL">num_dots</div>
<div class="feature-chip" data-tip="Number of hyphens (-) in the full URL">num_hyphens</div>
<div class="feature-chip" data-tip="Number of underscores (_) in the full URL">num_underscores</div>
<div class="feature-chip" data-tip="Number of forward slashes (/) in the URL">num_slashes</div>
<div class="feature-chip" data-tip="Number of question marks (?) in the URL">num_question_marks</div>
<div class="feature-chip" data-tip="Number of ampersands (&) in the URL">num_ampersands</div>
<div class="feature-chip" data-tip="Number of equals signs (=) in the URL">num_equals</div>
<div class="feature-chip" data-tip="Number of @ symbols — often used to obscure the real destination">num_at</div>
<div class="feature-chip" data-tip="Number of percent (%) characters indicating URL encoding">num_percent</div>
<div class="feature-chip" data-tip="Total digit characters in the full URL">num_digits_url</div>
<div class="feature-chip" data-tip="Total letter characters in the full URL">num_letters_url</div>
<div class="feature-chip" data-tip="Number of dots (.) in the domain only">domain_dots</div>
<div class="feature-chip" data-tip="Number of hyphens (-) in the domain only">domain_hyphens</div>
<div class="feature-chip" data-tip="Number of digit characters in the domain">domain_digits</div>
<div class="feature-chip" data-tip="Number of slashes (/) in the path component">path_slashes</div>
<div class="feature-chip" data-tip="Number of dots (.) in the path component">path_dots</div>
<div class="feature-chip" data-tip="Number of digit characters in the path">path_digits</div>
</div>
<div class="feature-category-label">Character Ratios</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Proportion of digit characters to total URL length">digit_ratio_url</div>
<div class="feature-chip" data-tip="Proportion of letter characters to total URL length">letter_ratio_url</div>
<div class="feature-chip" data-tip="Proportion of special (non-alphanumeric) characters in URL">special_char_ratio</div>
<div class="feature-chip" data-tip="Proportion of digits in the domain name">digit_ratio_domain</div>
<div class="feature-chip" data-tip="Proportion of symbols (hyphens, underscores, dots) in domain">symbol_ratio_domain</div>
</div>
<div class="feature-category-label">Domain Structure</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Number of subdomains (e.g. sub.example.com = 1)">num_subdomains</div>
<div class="feature-chip" data-tip="Total number of dot-separated domain parts">num_domain_parts</div>
<div class="feature-chip" data-tip="Character length of the top-level domain (e.g. com=3)">tld_length</div>
<div class="feature-chip" data-tip="Character length of the second-level domain">sld_length</div>
<div class="feature-chip" data-tip="Length of the longest dot-separated domain part">longest_domain_part</div>
<div class="feature-chip" data-tip="Average length of all domain parts">avg_domain_part_len</div>
<div class="feature-chip" data-tip="1 if any domain part exceeds 20 characters">longest_part_gt_20</div>
<div class="feature-chip" data-tip="1 if any domain part exceeds 30 characters">longest_part_gt_30</div>
<div class="feature-chip" data-tip="1 if any domain part exceeds 40 characters">longest_part_gt_40</div>
<div class="feature-chip" data-tip="1 if TLD is suspicious (.tk, .ml, .xyz, .top, .zip, etc.)">has_suspicious_tld</div>
<div class="feature-chip" data-tip="1 if TLD is well-known and trusted (.com, .org, .edu, etc.)">has_trusted_tld</div>
<div class="feature-chip" data-tip="1 if URL contains a port number">has_port</div>
<div class="feature-chip" data-tip="1 if URL uses a non-standard port (not 80 or 443)">has_non_std_port</div>
<div class="feature-chip" data-tip="Composite randomness score of the domain (0-1)">domain_randomness_score</div>
<div class="feature-chip" data-tip="Consonant clustering score of the SLD — random strings have high clusters">sld_consonant_cluster_score</div>
<div class="feature-chip" data-tip="1 if SLD contains keyboard walk patterns (qwerty, asdfgh)">sld_keyboard_pattern</div>
<div class="feature-chip" data-tip="1 if SLD contains a common English word (4+ characters)">sld_has_dictionary_word</div>
<div class="feature-chip" data-tip="Score based on vowel/consonant alternation — real words are more pronounceable">sld_pronounceability_score</div>
<div class="feature-chip" data-tip="1 if digits appear at suspicious positions in the SLD (start or end)">domain_digit_position_suspicious</div>
</div>
<div class="feature-category-label">Path Analysis</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Directory depth of the URL path (number of segments)">path_depth</div>
<div class="feature-chip" data-tip="Length of the longest path segment between slashes">max_path_segment_len</div>
<div class="feature-chip" data-tip="Average length of path segments">avg_path_segment_len</div>
<div class="feature-chip" data-tip="1 if the URL path has a file extension">has_extension</div>
<div class="feature-chip" data-tip="Category of file extension: 0=none, 1=document, 2=media, 3=executable, 4=web, 5=other">extension_category</div>
<div class="feature-chip" data-tip="1 if extension is suspicious (.exe, .bat, .cmd, .scr, .vbs, .ps1)">has_suspicious_extension</div>
<div class="feature-chip" data-tip="1 if extension is specifically .exe">has_exe</div>
<div class="feature-chip" data-tip="1 if path contains double slash (//) — possible redirect trick">has_double_slash</div>
<div class="feature-chip" data-tip="1 if a brand name appears in path but not in domain — impersonation signal">path_has_brand_not_domain</div>
<div class="feature-chip" data-tip="1 if path contains an IP address pattern">path_has_ip_pattern</div>
<div class="feature-chip" data-tip="1 if a document extension + 'download' keyword appear together">suspicious_path_extension_combo</div>
</div>
<div class="feature-category-label">Query String</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Number of query parameters">num_params</div>
<div class="feature-chip" data-tip="1 if URL has a query string">has_query</div>
<div class="feature-chip" data-tip="Total character length of all query parameter values">query_value_length</div>
<div class="feature-chip" data-tip="Length of the longest query parameter">max_param_len</div>
<div class="feature-chip" data-tip="1 if any query parameter value looks like a URL — possible redirect">query_has_url</div>
</div>
<div class="feature-category-label">Statistical & Entropy</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Shannon entropy of the full URL — random/phishing URLs have higher entropy">url_entropy</div>
<div class="feature-chip" data-tip="Shannon entropy of the domain name">domain_entropy</div>
<div class="feature-chip" data-tip="Shannon entropy of the path component">path_entropy</div>
<div class="feature-chip" data-tip="Longest run of consecutive digit characters">max_consecutive_digits</div>
<div class="feature-chip" data-tip="Longest run of consecutive letter characters">max_consecutive_chars</div>
<div class="feature-chip" data-tip="Longest run of consecutive consonants in domain">max_consecutive_consonants</div>
<div class="feature-chip" data-tip="Rate of adjacent character repetitions (aa, bb, etc.)">char_repeat_rate</div>
<div class="feature-chip" data-tip="Ratio of unique bigrams — lower = more repetitive URL">unique_bigram_ratio</div>
<div class="feature-chip" data-tip="Ratio of unique trigrams — lower = more repetitive URL">unique_trigram_ratio</div>
<div class="feature-chip" data-tip="Ratio of unique characters to total characters in the SLD (0–1)">sld_letter_diversity</div>
<div class="feature-chip" data-tip="1 if domain contains both digits and letters">domain_has_numbers_letters</div>
<div class="feature-chip" data-tip="Composite score (0–1) combining URL length, dots, hyphens, slashes, and entropy">url_complexity_score</div>
</div>
<div class="feature-category-label">Security Indicators</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="1 if domain is an IP address instead of a hostname">has_ip_address</div>
<div class="feature-chip" data-tip="1 if URL contains @ symbol — can trick browsers into treating text before @ as user info">has_at_symbol</div>
<div class="feature-chip" data-tip="1 if double-slash redirect pattern found in path">has_redirect</div>
<div class="feature-chip" data-tip="1 if domain is a known URL shortener (bit.ly, t.co, etc.)">is_shortened</div>
<div class="feature-chip" data-tip="1 if hosted on a free hosting service (000webhostapp, freehosting, etc.)">is_free_hosting</div>
<div class="feature-chip" data-tip="1 if hosted on a free platform (github.io, vercel.app, netlify.app, etc.)">is_free_platform</div>
<div class="feature-chip" data-tip="Length of subdomain on free platforms — long random subdomains are suspicious">platform_subdomain_length</div>
<div class="feature-chip" data-tip="1 if subdomain matches UUID-like pattern (common on Replit, Firebase)">has_uuid_subdomain</div>
<div class="feature-chip" data-tip="1 if URL uses HTTP instead of HTTPS">is_http</div>
</div>
<div class="feature-category-label">Keywords & Brand Detection</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Count of phishing keywords (login, verify, secure, etc.) in URL">num_phishing_keywords</div>
<div class="feature-chip" data-tip="1 if phishing keywords found in the domain">phishing_in_domain</div>
<div class="feature-chip" data-tip="1 if phishing keywords found in the path">phishing_in_path</div>
<div class="feature-chip" data-tip="Number of recognized brand names in the URL">num_brands</div>
<div class="feature-chip" data-tip="1 if a brand name appears in the domain">brand_in_domain</div>
<div class="feature-chip" data-tip="1 if a brand name appears in the path">brand_in_path</div>
<div class="feature-chip" data-tip="Score for brand impersonation — brand in URL but not the real domain">brand_impersonation</div>
<div class="feature-chip" data-tip="1 if 'login' appears in URL">has_login</div>
<div class="feature-chip" data-tip="1 if 'account' appears in URL">has_account</div>
<div class="feature-chip" data-tip="1 if 'verify' appears in URL">has_verify</div>
<div class="feature-chip" data-tip="1 if 'secure' appears in URL">has_secure</div>
<div class="feature-chip" data-tip="1 if 'update' appears in URL">has_update</div>
<div class="feature-chip" data-tip="1 if 'bank' appears in URL">has_bank</div>
<div class="feature-chip" data-tip="1 if 'password' or 'passwd' appears in URL">has_password</div>
<div class="feature-chip" data-tip="1 if 'suspend' appears in URL">has_suspend</div>
<div class="feature-chip" data-tip="1 if 'webscr' appears in URL — common in PayPal phishing">has_webscr</div>
<div class="feature-chip" data-tip="1 if 'cmd=' or '/cmd/' appears in URL">has_cmd</div>
<div class="feature-chip" data-tip="1 if 'cgi-bin' or '.cgi' appears in URL">has_cgi</div>
<div class="feature-chip" data-tip="1 if brand name in subdomain but not main domain — spoofing pattern">brand_in_subdomain_not_domain</div>
<div class="feature-chip" data-tip="1 if multiple different brand names detected in URL">multiple_brands_in_url</div>
<div class="feature-chip" data-tip="1 if brand name combined with hyphen (e.g. paypal-login.com)">brand_with_hyphen</div>
<div class="feature-chip" data-tip="1 if brand found in domain with suspicious TLD">suspicious_brand_tld</div>
<div class="feature-chip" data-tip="1 if brand name + phishing keyword both present">brand_keyword_combo</div>
</div>
<div class="feature-category-label">Encoding & Obfuscation</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="1 if URL contains percent-encoded characters">has_url_encoding</div>
<div class="feature-chip" data-tip="Number of percent-encoded sequences in URL">encoding_count</div>
<div class="feature-chip" data-tip="Difference in length between encoded and decoded URL">encoding_diff</div>
<div class="feature-chip" data-tip="1 if domain contains Punycode (xn-- prefix) — internationalized domain">has_punycode</div>
<div class="feature-chip" data-tip="1 if URL contains non-ASCII Unicode characters">has_unicode</div>
<div class="feature-chip" data-tip="1 if URL contains hexadecimal string (0x...)">has_hex_string</div>
<div class="feature-chip" data-tip="1 if URL contains a Base64-like string (20+ alphanumeric chars with +/=)">has_base64</div>
<div class="feature-chip" data-tip="1 if domain contains look-alike patterns (rn, vv, cl, 0, 1) that mimic other characters">has_lookalike_chars</div>
<div class="feature-chip" data-tip="Score for mixed Unicode scripts in domain — homograph attack indicator">mixed_script_score</div>
<div class="feature-chip" data-tip="Risk score for homograph attacks targeting brand names">homograph_brand_risk</div>
<div class="feature-chip" data-tip="1 if IDN homograph score exceeds 0.5 threshold">suspected_idn_homograph</div>
<div class="feature-chip" data-tip="1 if URL contains double percent-encoding (%% or %25)">double_encoding</div>
<div class="feature-chip" data-tip="1 if percent-encoding found specifically in the domain">encoding_in_domain</div>
<div class="feature-chip" data-tip="Count of suspicious Unicode characters (RTL override, zero-width, BOM), capped at 5">suspicious_unicode_category</div>
</div>
</div>
</section>
<!-- HTML FEATURES -->
<section class="section">
<div class="section-title collapsible-toggle" onclick="toggleFeatures(this)">
RAW HTML Features <span class="feature-count">77 raw features + 23 engineered</span>
<span class="toggle-icon">+</span>
</div>
<div class="collapsible-content">
<div class="section-subtitle">All features extracted from HTML source and DOM structure. Hover over any feature to see its description.</div>
<div class="feature-category-label">Document Size & Text</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Maximum nesting depth of DOM elements">dom_depth</div>
<div class="feature-chip" data-tip="Total character length of the raw HTML">html_length</div>
<div class="feature-chip" data-tip="Total length of visible extracted text">text_length</div>
<div class="feature-chip" data-tip="Number of words extracted from page text">num_words</div>
<div class="feature-chip" data-tip="Ratio of text content length to full HTML length">text_to_html_ratio</div>
<div class="feature-chip" data-tip="Total character length of inline CSS styles">inline_css_length</div>
<div class="feature-chip" data-tip="Total number of HTML tags in the document">num_tags</div>
</div>
<div class="feature-category-label">Metadata & Page Identity</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="1 if HTML title tag is present">has_title</div>
<div class="feature-chip" data-tip="1 if meta description tag is present">has_description</div>
<div class="feature-chip" data-tip="1 if meta keywords tag is present">has_keywords</div>
<div class="feature-chip" data-tip="1 if author metadata is present">has_author</div>
<div class="feature-chip" data-tip="1 if copyright text or metadata is detected">has_copyright</div>
<div class="feature-chip" data-tip="1 if viewport meta tag is present">has_viewport</div>
<div class="feature-chip" data-tip="1 if favicon link is declared">has_favicon</div>
<div class="feature-chip" data-tip="Number of meta tags in the page head">num_meta_tags</div>
</div>
<div class="feature-category-label">DOM Elements & Layout</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Number of div elements">num_divs</div>
<div class="feature-chip" data-tip="Number of span elements">num_spans</div>
<div class="feature-chip" data-tip="Number of paragraph tags">num_paragraphs</div>
<div class="feature-chip" data-tip="Number of heading tags (h1-h6)">num_headings</div>
<div class="feature-chip" data-tip="Number of list containers (ul, ol)">num_lists</div>
<div class="feature-chip" data-tip="Number of table elements">num_tables</div>
<div class="feature-chip" data-tip="Number of image elements">num_images</div>
<div class="feature-chip" data-tip="Number of iframe elements">num_iframes</div>
<div class="feature-chip" data-tip="Number of hidden iframes">num_hidden_iframes</div>
<div class="feature-chip" data-tip="Number of images embedded as data URIs">num_data_uri_images</div>
<div class="feature-chip" data-tip="Number of linked CSS files">num_css_files</div>
<div class="feature-chip" data-tip="Number of script tags">num_scripts</div>
<div class="feature-chip" data-tip="Number of inline script blocks">num_inline_scripts</div>
<div class="feature-chip" data-tip="Number of inline style blocks or style attributes">num_inline_styles</div>
<div class="feature-chip" data-tip="Number of input fields in forms">num_input_fields</div>
</div>
<div class="feature-category-label">Link & Resource Analysis</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Total number of links">num_links</div>
<div class="feature-chip" data-tip="Number of internal links pointing to same domain">num_internal_links</div>
<div class="feature-chip" data-tip="Number of external links pointing to other domains">num_external_links</div>
<div class="feature-chip" data-tip="Ratio of external links to all links">ratio_external_links</div>
<div class="feature-chip" data-tip="Count of distinct external domains referenced">num_unique_external_domains</div>
<div class="feature-chip" data-tip="Number of mailto links">num_mailto_links</div>
<div class="feature-chip" data-tip="Number of javascript: pseudo-links">num_javascript_links</div>
<div class="feature-chip" data-tip="Number of links pointing to IP-based hosts">num_ip_based_links</div>
<div class="feature-chip" data-tip="Number of links using suspicious top-level domains">num_suspicious_tld_links</div>
<div class="feature-chip" data-tip="Number of links with empty or placeholder href values">num_empty_links</div>
<div class="feature-chip" data-tip="Anchor text points to content unrelated to destination">num_anchor_text_mismatch</div>
<div class="feature-chip" data-tip="Number of external stylesheet references">num_external_css</div>
<div class="feature-chip" data-tip="Number of externally loaded images">num_external_images</div>
<div class="feature-chip" data-tip="Number of externally loaded JavaScript files">num_external_scripts</div>
</div>
<div class="feature-category-label">Forms & Inputs</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="1 if at least one form tag exists">has_form</div>
<div class="feature-chip" data-tip="1 if a login-style form is detected">has_login_form</div>
<div class="feature-chip" data-tip="Number of form elements">num_forms</div>
<div class="feature-chip" data-tip="Number of email-type fields">num_email_fields</div>
<div class="feature-chip" data-tip="Number of password fields">num_password_fields</div>
<div class="feature-chip" data-tip="Number of text input fields">num_text_fields</div>
<div class="feature-chip" data-tip="Number of submit buttons">num_submit_buttons</div>
<div class="feature-chip" data-tip="Number of hidden input fields">num_hidden_fields</div>
<div class="feature-chip" data-tip="Number of forms missing associated labels">num_forms_without_labels</div>
<div class="feature-chip" data-tip="Number of forms with empty action attribute">num_empty_form_actions</div>
<div class="feature-chip" data-tip="Number of forms submitting to external domains">num_external_form_actions</div>
<div class="feature-chip" data-tip="1 if password form submits to external domain">password_with_external_action</div>
</div>
<div class="feature-category-label">Scripts & Dynamic Behavior</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="1 if JavaScript eval() is used">has_eval</div>
<div class="feature-chip" data-tip="1 if JavaScript escape() is used">has_escape</div>
<div class="feature-chip" data-tip="1 if JavaScript unescape() is used">has_unescape</div>
<div class="feature-chip" data-tip="1 if atob() decoding function is present">has_atob</div>
<div class="feature-chip" data-tip="1 if Base64-like content or decoding usage is detected">has_base64</div>
<div class="feature-chip" data-tip="1 if String.fromCharCode usage is present">has_fromcharcode</div>
<div class="feature-chip" data-tip="1 if document.write() is used">has_document_write</div>
<div class="feature-chip" data-tip="1 if window.open() is used">has_window_open</div>
<div class="feature-chip" data-tip="1 if location.replace() redirects are used">has_location_replace</div>
<div class="feature-chip" data-tip="1 if meta refresh redirect is present">has_meta_refresh</div>
<div class="feature-chip" data-tip="Number of onclick event handlers">num_onclick_events</div>
<div class="feature-chip" data-tip="Number of onload event handlers">num_onload_events</div>
<div class="feature-chip" data-tip="Number of onerror event handlers">num_onerror_events</div>
</div>
<div class="feature-category-label">Visibility & Interaction Tricks</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="1 if display:none usage is detected">has_display_none</div>
<div class="feature-chip" data-tip="1 if visibility:hidden usage is detected">has_visibility_hidden</div>
<div class="feature-chip" data-tip="1 if right-click is disabled by script">has_right_click_disabled</div>
<div class="feature-chip" data-tip="1 if status bar text customization is attempted">has_status_bar_customization</div>
</div>
<div class="feature-category-label">Contact & Social Engineering Signals</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="1 if raw email address patterns appear in HTML">has_email_address</div>
<div class="feature-chip" data-tip="1 if phone number patterns appear in HTML">has_phone_number</div>
<div class="feature-chip" data-tip="Number of known brand name mentions in HTML text">num_brand_mentions</div>
<div class="feature-chip" data-tip="Number of urgency words (urgent, verify, immediately, etc.)">num_urgency_keywords</div>
</div>
<div class="feature-category-label">Ratios & Proportions</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Number of forms divided by number of input fields — phishing sites often have few forms with many inputs">forms_to_inputs_ratio</div>
<div class="feature-chip" data-tip="Proportion of external links to total links — high ratio indicates off-site redirection">external_to_total_links</div>
<div class="feature-chip" data-tip="Number of scripts divided by total HTML tags — high ratio suggests heavy JavaScript reliance">scripts_to_tags_ratio</div>
<div class="feature-chip" data-tip="Ratio of hidden to visible input fields — hidden fields often used for tracking or obfuscation">hidden_to_visible_inputs</div>
<div class="feature-chip" data-tip="Number of password fields divided by total input fields — phishing forms maximize password collection">password_to_inputs_ratio</div>
<div class="feature-chip" data-tip="Proportion of empty/placeholder links to all links — broken or disguised navigation is suspicious">empty_to_total_links</div>
<div class="feature-chip" data-tip="Number of images divided by total HTML tags — low image ratio suggests text-heavy phishing pages">images_to_tags_ratio</div>
<div class="feature-chip" data-tip="Number of iframes divided by total HTML tags — iframes can hide malicious content or redirects">iframes_to_tags_ratio</div>
</div>
<div class="feature-category-label">Co-occurrence Interactions</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="1 if forms exist AND password fields are present — core phishing indicator">forms_with_passwords</div>
<div class="feature-chip" data-tip="Count of external JavaScript files — malicious JS often hosted externally">external_scripts_links</div>
<div class="feature-chip" data-tip="1 if urgency keywords AND forms both present — common social engineering pattern">urgency_with_forms</div>
<div class="feature-chip" data-tip="1 if brand names AND forms both present — brand impersonation with credential harvesting">brand_with_forms</div>
<div class="feature-chip" data-tip="1 if iframes AND script tags both present — hidden iframes + JS often for malware">iframes_with_scripts</div>
<div class="feature-chip" data-tip="1 if hidden inputs AND external resources both present — obfuscated tracking/redirects">hidden_with_external</div>
</div>
<div class="feature-category-label">Content Density</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Ratio of visible text length to DOM depth — legitimate sites have better content distribution">content_density</div>
<div class="feature-chip" data-tip="Number of form elements divided by total words — form-heavy pages are suspicious">form_density</div>
<div class="feature-chip" data-tip="Number of script tags divided by number of forms — scripts per form indicates obfuscation level">scripts_per_form</div>
<div class="feature-chip" data-tip="Number of links divided by total words — link density compared to content volume">links_per_word</div>
</div>
<div class="feature-category-label">Composite Risk Scores</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="Composite score (0–1) combining form presence, password fields, external links, and scripts">phishing_risk_score</div>
<div class="feature-chip" data-tip="Score (0–1) specifically for form-based threats: password fields, hidden inputs, external actions">form_risk_score</div>
<div class="feature-chip" data-tip="Score (0–1) measuring obfuscation techniques: hidden fields, eval usage, encoding, meta refreshes">obfuscation_score</div>
<div class="feature-chip" data-tip="Score (0–1) measuring legitimacy signals: metadata presence, proper structure, internal links">legitimacy_score</div>
</div>
<div class="feature-category-label">Boolean Flags</div>
<div class="feature-grid">
<div class="feature-chip" data-tip="1 if any combination of suspicious elements detected (hidden inputs + eval, eval + meta refresh, etc.)">has_suspicious_elements</div>
</div>
</div>
</section>
<!-- TABS: MODEL CATEGORIES -->
<section class="section" style="border-bottom:none;padding-bottom:0">
<div class="tabs">
<button class="tab active" onclick="switchTab(event,'urlModels')">URL Models <span class="tab-count">3</span></button>
<button class="tab" onclick="switchTab(event,'htmlModels')">HTML Models <span class="tab-count">2</span></button>
<button class="tab" onclick="switchTab(event,'combinedModels')">Combined <span class="tab-count">2</span></button>
<button class="tab" onclick="switchTab(event,'cnnModels')">CNN <span class="tab-count">2</span></button>
<button class="tab" onclick="switchTab(event,'overview')">Comparison</button>
</div>
<!-- URL MODELS TAB -->
<div id="urlModels" class="tab-content active">
<div class="section-subtitle">
3 models trained on 120 URL-based features extracted from the URL string structure,
domain properties, encoding analysis, and brand impersonation detection.
</div>
<!-- Logistic Regression -->
<div class="model-detail">
<div class="model-detail-header">
<div class="model-detail-name">Logistic Regression</div>
<div class="model-detail-type">Baseline</div>
</div>
<div class="metrics-grid">
<div class="metric-card">
<div class="metric-value">93.71%</div>
<div class="metric-label">Accuracy</div>
</div>
<div class="metric-card">
<div class="metric-value">95.40%</div>
<div class="metric-label">Precision</div>
</div>
<div class="metric-card">
<div class="metric-value">91.84%</div>
<div class="metric-label">Recall</div>
</div>
<div class="metric-card">
<div class="metric-value">93.59%</div>
<div class="metric-label">F1-Score</div>
</div>
<div class="metric-card">
<div class="metric-value highlight">0.9789</div>
<div class="metric-label">ROC-AUC</div>
</div>
</div>
<div class="section-title">Confusion Matrix</div>
<div class="confusion-matrix">
<div class="cm-header"></div>
<div class="cm-header">Pred Legit</div>
<div class="cm-header">Pred Phish</div>
<div class="cm-label">Actual Legit</div>
<div class="cm-cell cm-tn">10,326</div>
<div class="cm-cell cm-fp">478</div>
<div class="cm-label">Actual Phish</div>
<div class="cm-cell cm-fn">881</div>
<div class="cm-cell cm-tp">9,922</div>
</div>
</div>
<!-- Random Forest -->
<div class="model-detail">
<div class="model-detail-header">
<div class="model-detail-name">Random Forest</div>
<div class="model-detail-type">Ensemble</div>
</div>
<div class="metrics-grid">
<div class="metric-card">
<div class="metric-value">97.63%</div>
<div class="metric-label">Accuracy</div>
</div>
<div class="metric-card">
<div class="metric-value">99.01%</div>
<div class="metric-label">Precision</div>
</div>
<div class="metric-card">
<div class="metric-value">96.22%</div>
<div class="metric-label">Recall</div>
</div>
<div class="metric-card">
<div class="metric-value">97.60%</div>
<div class="metric-label">F1-Score</div>
</div>
<div class="metric-card">
<div class="metric-value highlight">0.9958</div>
<div class="metric-label">ROC-AUC</div>
</div>
</div>
<div class="section-title">Confusion Matrix</div>
<div class="confusion-matrix">
<div class="cm-header"></div>
<div class="cm-header">Pred Legit</div>
<div class="cm-header">Pred Phish</div>
<div class="cm-label">Actual Legit</div>
<div class="cm-cell cm-tn">10,700</div>
<div class="cm-cell cm-fp">104</div>
<div class="cm-label">Actual Phish</div>
<div class="cm-cell cm-fn">408</div>
<div class="cm-cell cm-tp">10,395</div>
</div>
<div class="subsection">
<div class="section-title">Top 20 Features by Importance</div>
<div class="features-list">
<div class="feature-row"><span class="feature-rank">1</span><span class="feature-name">domain_length</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:100%"></div></div><span class="feature-importance">0.0500</span></div>
<div class="feature-row"><span class="feature-rank">2</span><span class="feature-name">num_domain_parts</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:94%"></div></div><span class="feature-importance">0.0471</span></div>
<div class="feature-row"><span class="feature-rank">3</span><span class="feature-name">domain_dots</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:91%"></div></div><span class="feature-importance">0.0453</span></div>
<div class="feature-row"><span class="feature-rank">4</span><span class="feature-name">num_subdomains</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:79%"></div></div><span class="feature-importance">0.0393</span></div>
<div class="feature-row"><span class="feature-rank">5</span><span class="feature-name">num_dots</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:67%"></div></div><span class="feature-importance">0.0337</span></div>
<div class="feature-row"><span class="feature-rank">6</span><span class="feature-name">domain_length_category</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:67%"></div></div><span class="feature-importance">0.0335</span></div>
<div class="feature-row"><span class="feature-rank">7</span><span class="feature-name">symbol_ratio_domain</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:65%"></div></div><span class="feature-importance">0.0324</span></div>
<div class="feature-row"><span class="feature-rank">8</span><span class="feature-name">digit_ratio_url</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:60%"></div></div><span class="feature-importance">0.0298</span></div>
<div class="feature-row"><span class="feature-rank">9</span><span class="feature-name">path_length</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:59%"></div></div><span class="feature-importance">0.0297</span></div>
<div class="feature-row"><span class="feature-rank">10</span><span class="feature-name">avg_domain_part_len</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:58%"></div></div><span class="feature-importance">0.0292</span></div>
<div class="feature-row"><span class="feature-rank">11</span><span class="feature-name">num_digits_url</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:57%"></div></div><span class="feature-importance">0.0283</span></div>
<div class="feature-row"><span class="feature-rank">12</span><span class="feature-name">domain_entropy</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:56%"></div></div><span class="feature-importance">0.0282</span></div>
<div class="feature-row"><span class="feature-rank">13</span><span class="feature-name">url_entropy</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:53%"></div></div><span class="feature-importance">0.0266</span></div>
<div class="feature-row"><span class="feature-rank">14</span><span class="feature-name">max_consecutive_digits</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:49%"></div></div><span class="feature-importance">0.0246</span></div>
<div class="feature-row"><span class="feature-rank">15</span><span class="feature-name">special_char_ratio</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:48%"></div></div><span class="feature-importance">0.0242</span></div>
<div class="feature-row"><span class="feature-rank">16</span><span class="feature-name">is_shortened</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:47%"></div></div><span class="feature-importance">0.0237</span></div>
<div class="feature-row"><span class="feature-rank">17</span><span class="feature-name">path_entropy</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:45%"></div></div><span class="feature-importance">0.0225</span></div>
<div class="feature-row"><span class="feature-rank">18</span><span class="feature-name">max_path_segment_len</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:43%"></div></div><span class="feature-importance">0.0215</span></div>
<div class="feature-row"><span class="feature-rank">19</span><span class="feature-name">num_letters_url</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:41%"></div></div><span class="feature-importance">0.0206</span></div>
<div class="feature-row"><span class="feature-rank">20</span><span class="feature-name">path_slashes</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:40%"></div></div><span class="feature-importance">0.0201</span></div>
</div>
</div>
</div>
<!-- XGBoost -->
<div class="model-detail">
<div class="model-detail-header">
<div class="model-detail-name">XGBoost</div>
<div class="model-detail-type">Gradient Boosting</div>
</div>
<div class="metrics-grid">
<div class="metric-card">
<div class="metric-value">97.85%</div>
<div class="metric-label">Accuracy</div>
</div>
<div class="metric-card">
<div class="metric-value">99.00%</div>
<div class="metric-label">Precision</div>
</div>
<div class="metric-card">
<div class="metric-value">96.68%</div>
<div class="metric-label">Recall</div>
</div>
<div class="metric-card">
<div class="metric-value">97.82%</div>
<div class="metric-label">F1-Score</div>
</div>
<div class="metric-card">
<div class="metric-value highlight">0.9953</div>
<div class="metric-label">ROC-AUC</div>
</div>
</div>
<div class="section-title">Confusion Matrix</div>
<div class="confusion-matrix">
<div class="cm-header"></div>
<div class="cm-header">Pred Legit</div>
<div class="cm-header">Pred Phish</div>
<div class="cm-label">Actual Legit</div>
<div class="cm-cell cm-tn">10,698</div>
<div class="cm-cell cm-fp">106</div>
<div class="cm-label">Actual Phish</div>
<div class="cm-cell cm-fn">359</div>
<div class="cm-cell cm-tp">10,444</div>
</div>
<div class="subsection">
<div class="section-title">Top 20 Features by Importance</div>
<div class="features-list">
<div class="feature-row"><span class="feature-rank">1</span><span class="feature-name">domain_dots</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:100%"></div></div><span class="feature-importance">0.2514</span></div>
<div class="feature-row"><span class="feature-rank">2</span><span class="feature-name">is_shortened</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:60%"></div></div><span class="feature-importance">0.1519</span></div>
<div class="feature-row"><span class="feature-rank">3</span><span class="feature-name">num_subdomains</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:57%"></div></div><span class="feature-importance">0.1423</span></div>
<div class="feature-row"><span class="feature-rank">4</span><span class="feature-name">num_domain_parts</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:20%"></div></div><span class="feature-importance">0.0492</span></div>
<div class="feature-row"><span class="feature-rank">5</span><span class="feature-name">multiple_brands_in_url</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:14%"></div></div><span class="feature-importance">0.0350</span></div>
<div class="feature-row"><span class="feature-rank">6</span><span class="feature-name">is_free_platform</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:11%"></div></div><span class="feature-importance">0.0281</span></div>
<div class="feature-row"><span class="feature-rank">7</span><span class="feature-name">domain_hyphens</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:10%"></div></div><span class="feature-importance">0.0252</span></div>
<div class="feature-row"><span class="feature-rank">8</span><span class="feature-name">path_digits</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:6%"></div></div><span class="feature-importance">0.0149</span></div>
<div class="feature-row"><span class="feature-rank">9</span><span class="feature-name">is_http</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:6%"></div></div><span class="feature-importance">0.0139</span></div>
<div class="feature-row"><span class="feature-rank">10</span><span class="feature-name">platform_subdomain_length</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:5%"></div></div><span class="feature-importance">0.0123</span></div>
<div class="feature-row"><span class="feature-rank">11</span><span class="feature-name">avg_domain_part_len</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:5%"></div></div><span class="feature-importance">0.0121</span></div>
<div class="feature-row"><span class="feature-rank">12</span><span class="feature-name">path_slashes</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:5%"></div></div><span class="feature-importance">0.0119</span></div>
<div class="feature-row"><span class="feature-rank">13</span><span class="feature-name">brand_in_path</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:4%"></div></div><span class="feature-importance">0.0111</span></div>
<div class="feature-row"><span class="feature-rank">14</span><span class="feature-name">domain_length_category</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:4%"></div></div><span class="feature-importance">0.0108</span></div>
<div class="feature-row"><span class="feature-rank">15</span><span class="feature-name">domain_length</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:4%"></div></div><span class="feature-importance">0.0106</span></div>
<div class="feature-row"><span class="feature-rank">16</span><span class="feature-name">symbol_ratio_domain</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:4%"></div></div><span class="feature-importance">0.0100</span></div>
<div class="feature-row"><span class="feature-rank">17</span><span class="feature-name">encoding_diff</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:3%"></div></div><span class="feature-importance">0.0087</span></div>
<div class="feature-row"><span class="feature-rank">18</span><span class="feature-name">num_brands</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:3%"></div></div><span class="feature-importance">0.0077</span></div>
<div class="feature-row"><span class="feature-rank">19</span><span class="feature-name">tld_length</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:3%"></div></div><span class="feature-importance">0.0071</span></div>
<div class="feature-row"><span class="feature-rank">20</span><span class="feature-name">digit_ratio_url</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:3%"></div></div><span class="feature-importance">0.0067</span></div>
</div>
</div>
</div>
</div>
<!-- HTML MODELS TAB -->
<div id="htmlModels" class="tab-content">
<div class="section-subtitle">
2 models trained on 100 HTML-based features extracted from the page structure,
forms, scripts, links, and content analysis of downloaded web pages.
</div>
<!-- Random Forest HTML -->
<div class="model-detail">
<div class="model-detail-header">
<div class="model-detail-name">Random Forest HTML</div>
<div class="model-detail-type">Ensemble</div>
</div>
<div class="metrics-grid">
<div class="metric-card">
<div class="metric-value">89.65%</div>
<div class="metric-label">Accuracy</div>
</div>
<div class="metric-card">
<div class="metric-value">91.78%</div>
<div class="metric-label">Precision</div>
</div>
<div class="metric-card">
<div class="metric-value">87.11%</div>
<div class="metric-label">Recall</div>
</div>
<div class="metric-card">
<div class="metric-value">89.38%</div>
<div class="metric-label">F1-Score</div>
</div>
<div class="metric-card">
<div class="metric-value highlight">0.9617</div>
<div class="metric-label">ROC-AUC</div>
</div>
<div class="metric-card">
<div class="metric-value">89.05%</div>
<div class="metric-label">CV F1 (5-fold)</div>
</div>
</div>
<div class="section-title">Confusion Matrix</div>
<div class="confusion-matrix">
<div class="cm-header"></div>
<div class="cm-header">Pred Legit</div>
<div class="cm-header">Pred Phish</div>
<div class="cm-label">Actual Legit</div>
<div class="cm-cell cm-tn">15,012</div>
<div class="cm-cell cm-fp">1,271</div>
<div class="cm-label">Actual Phish</div>
<div class="cm-cell cm-fn">2,099</div>
<div class="cm-cell cm-tp">14,184</div>
</div>
<div class="subsection">
<div class="section-title">Hyperparameters</div>
<div class="params-grid">
<div class="param-item"><span class="param-key">n_estimators</span><span class="param-value">500</span></div>
<div class="param-item"><span class="param-key">max_depth</span><span class="param-value">35</span></div>
<div class="param-item"><span class="param-key">min_samples_split</span><span class="param-value">2</span></div>
<div class="param-item"><span class="param-key">min_samples_leaf</span><span class="param-value">1</span></div>
<div class="param-item"><span class="param-key">max_features</span><span class="param-value">sqrt</span></div>
<div class="param-item"><span class="param-key">class_weight</span><span class="param-value">balanced</span></div>
</div>
</div>
</div>
<!-- XGBoost HTML -->
<div class="model-detail">
<div class="model-detail-header">
<div class="model-detail-name">XGBoost HTML</div>
<div class="model-detail-type">Gradient Boosting</div>
</div>
<div class="metrics-grid">
<div class="metric-card">
<div class="metric-value">89.07%</div>
<div class="metric-label">Accuracy</div>
</div>
<div class="metric-card">
<div class="metric-value">90.27%</div>
<div class="metric-label">Precision</div>
</div>
<div class="metric-card">
<div class="metric-value">87.56%</div>
<div class="metric-label">Recall</div>
</div>
<div class="metric-card">
<div class="metric-value">88.90%</div>
<div class="metric-label">F1-Score</div>
</div>
<div class="metric-card">
<div class="metric-value highlight">0.9590</div>
<div class="metric-label">ROC-AUC</div>
</div>
<div class="metric-card">
<div class="metric-value">88.87%</div>
<div class="metric-label">CV F1 (5-fold)</div>
</div>
</div>
<div class="section-title">Confusion Matrix</div>
<div class="confusion-matrix">
<div class="cm-header"></div>
<div class="cm-header">Pred Legit</div>
<div class="cm-header">Pred Phish</div>
<div class="cm-label">Actual Legit</div>
<div class="cm-cell cm-tn">14,747</div>
<div class="cm-cell cm-fp">1,536</div>
<div class="cm-label">Actual Phish</div>
<div class="cm-cell cm-fn">2,025</div>
<div class="cm-cell cm-tp">14,258</div>
</div>
<div class="subsection">
<div class="section-title">Hyperparameters</div>
<div class="params-grid">
<div class="param-item"><span class="param-key">n_estimators</span><span class="param-value">600</span></div>
<div class="param-item"><span class="param-key">max_depth</span><span class="param-value">8</span></div>
<div class="param-item"><span class="param-key">learning_rate</span><span class="param-value">0.05</span></div>
<div class="param-item"><span class="param-key">subsample</span><span class="param-value">0.8</span></div>
<div class="param-item"><span class="param-key">colsample_bytree</span><span class="param-value">0.8</span></div>
<div class="param-item"><span class="param-key">min_child_weight</span><span class="param-value">3</span></div>
<div class="param-item"><span class="param-key">gamma</span><span class="param-value">0.1</span></div>
<div class="param-item"><span class="param-key">reg_alpha</span><span class="param-value">0.1</span></div>
<div class="param-item"><span class="param-key">reg_lambda</span><span class="param-value">1.0</span></div>
<div class="param-item"><span class="param-key">early_stopping</span><span class="param-value">50 rounds</span></div>
</div>
</div>
</div>
</div>
<!-- COMBINED MODELS TAB -->
<div id="combinedModels" class="tab-content">
<div class="section-subtitle">
2 models trained on 221 combined features (121 URL + 100 HTML) for maximum detection accuracy.
</div>
<!-- Random Forest Combined -->
<div class="model-detail">
<div class="model-detail-header">
<div class="model-detail-name">Random Forest Combined</div>
<div class="model-detail-type">Ensemble</div>
</div>
<div class="metrics-grid">
<div class="metric-card">
<div class="metric-value">98.60%</div>
<div class="metric-label">Accuracy</div>
</div>
<div class="metric-card">
<div class="metric-value">99.16%</div>
<div class="metric-label">Precision</div>
</div>
<div class="metric-card">
<div class="metric-value">98.02%</div>
<div class="metric-label">Recall</div>
</div>
<div class="metric-card">
<div class="metric-value">98.59%</div>
<div class="metric-label">F1-Score</div>
</div>
<div class="metric-card">
<div class="metric-value highlight">0.9990</div>
<div class="metric-label">ROC-AUC</div>
</div>
<div class="metric-card">
<div class="metric-value">98.59%</div>
<div class="metric-label">CV F1 (5-fold)</div>
</div>
</div>
<div class="section-title">Confusion Matrix</div>
<div class="confusion-matrix">
<div class="cm-header"></div>
<div class="cm-header">Pred Legit</div>
<div class="cm-header">Pred Phish</div>
<div class="cm-label">Actual Legit</div>
<div class="cm-cell cm-tn">10,680</div>
<div class="cm-cell cm-fp">89</div>
<div class="cm-label">Actual Phish</div>
<div class="cm-cell cm-fn">213</div>
<div class="cm-cell cm-tp">10,556</div>
</div>
<div class="subsection">
<div class="section-title">Feature Importance Split</div>
<div class="feature-split">
<div class="feature-split-bar">
<div class="split-url" style="width:29.1%">URL 29.1%</div>
<div class="split-html" style="width:70.9%">HTML 70.9%</div>
</div>
</div>
</div>
<div class="subsection">
<div class="section-title">Top 15 Features by Importance</div>
<div class="features-list">
<div class="feature-row"><span class="feature-rank">1</span><span class="feature-name">html_num_links</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:100%"></div></div><span class="feature-importance">0.0640</span></div>
<div class="feature-row"><span class="feature-rank">2</span><span class="feature-name">html_text_length</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:90%"></div></div><span class="feature-importance">0.0577</span></div>
<div class="feature-row"><span class="feature-rank">3</span><span class="feature-name">html_num_tags</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:75%"></div></div><span class="feature-importance">0.0479</span></div>
<div class="feature-row"><span class="feature-rank">4</span><span class="feature-name">html_num_internal_links</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:72%"></div></div><span class="feature-importance">0.0463</span></div>
<div class="feature-row"><span class="feature-rank">5</span><span class="feature-name">html_num_words</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:66%"></div></div><span class="feature-importance">0.0422</span></div>
<div class="feature-row"><span class="feature-rank">6</span><span class="feature-name">html_external_scripts_links</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:56%"></div></div><span class="feature-importance">0.0361</span></div>
<div class="feature-row"><span class="feature-rank">7</span><span class="feature-name">html_num_divs</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:46%"></div></div><span class="feature-importance">0.0297</span></div>
<div class="feature-row"><span class="feature-rank">8</span><span class="feature-name">html_num_lists</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:45%"></div></div><span class="feature-importance">0.0291</span></div>
<div class="feature-row"><span class="feature-rank">9</span><span class="feature-name">html_num_external_links</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:43%"></div></div><span class="feature-importance">0.0276</span></div>
<div class="feature-row"><span class="feature-rank">10</span><span class="feature-name">html_has_description</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:40%"></div></div><span class="feature-importance">0.0258</span></div>
<div class="feature-row"><span class="feature-rank">11</span><span class="feature-name">html_num_unique_external_domains</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:37%"></div></div><span class="feature-importance">0.0236</span></div>
<div class="feature-row"><span class="feature-rank">12</span><span class="feature-name">html_num_images</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:36%"></div></div><span class="feature-importance">0.0231</span></div>
<div class="feature-row"><span class="feature-rank">13</span><span class="feature-name">html_num_spans</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:35%"></div></div><span class="feature-importance">0.0226</span></div>
<div class="feature-row"><span class="feature-rank">14</span><span class="feature-name">html_num_headings</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:34%"></div></div><span class="feature-importance">0.0220</span></div>
<div class="feature-row"><span class="feature-rank">15</span><span class="feature-name">html_dom_depth</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:33%"></div></div><span class="feature-importance">0.0210</span></div>
</div>
</div>
<div class="subsection">
<div class="section-title">Hyperparameters</div>
<div class="params-grid">
<div class="param-item"><span class="param-key">n_estimators</span><span class="param-value">533</span></div>
<div class="param-item"><span class="param-key">max_depth</span><span class="param-value">43</span></div>
<div class="param-item"><span class="param-key">min_samples_split</span><span class="param-value">2</span></div>
<div class="param-item"><span class="param-key">max_features</span><span class="param-value">sqrt</span></div>
<div class="param-item"><span class="param-key">class_weight</span><span class="param-value">balanced</span></div>
</div>
</div>
</div>
<!-- XGBoost Combined -->
<div class="model-detail">
<div class="model-detail-header">
<div class="model-detail-name">XGBoost Combined</div>
<div class="model-detail-type">Gradient Boosting</div>
</div>
<div class="metrics-grid">
<div class="metric-card">
<div class="metric-value">99.01%</div>
<div class="metric-label">Accuracy</div>
</div>
<div class="metric-card">
<div class="metric-value">99.35%</div>
<div class="metric-label">Precision</div>
</div>
<div class="metric-card">
<div class="metric-value">98.66%</div>
<div class="metric-label">Recall</div>
</div>
<div class="metric-card">
<div class="metric-value">99.01%</div>
<div class="metric-label">F1-Score</div>
</div>
<div class="metric-card">
<div class="metric-value highlight">0.9991</div>
<div class="metric-label">ROC-AUC</div>
</div>
<div class="metric-card">
<div class="metric-value">98.90%</div>
<div class="metric-label">CV F1 (5-fold)</div>
</div>
</div>
<div class="section-title">Confusion Matrix</div>
<div class="confusion-matrix">
<div class="cm-header"></div>
<div class="cm-header">Pred Legit</div>
<div class="cm-header">Pred Phish</div>
<div class="cm-label">Actual Legit</div>
<div class="cm-cell cm-tn">10,700</div>
<div class="cm-cell cm-fp">69</div>
<div class="cm-label">Actual Phish</div>
<div class="cm-cell cm-fn">144</div>
<div class="cm-cell cm-tp">10,625</div>
</div>
<div class="subsection">
<div class="section-title">Feature Importance Split</div>
<div class="feature-split">
<div class="feature-split-bar">
<div class="split-url" style="width:37.1%">URL 37.1%</div>
<div class="split-html" style="width:62.9%">HTML 62.9%</div>
</div>
</div>
</div>
<div class="subsection">
<div class="section-title">Top 15 Features by Importance</div>
<div class="features-list">
<div class="feature-row"><span class="feature-rank">1</span><span class="feature-name">html_num_links</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:100%"></div></div><span class="feature-importance">0.4420</span></div>
<div class="feature-row"><span class="feature-rank">2</span><span class="feature-name">url_is_shortened</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:10%"></div></div><span class="feature-importance">0.0427</span></div>
<div class="feature-row"><span class="feature-rank">3</span><span class="feature-name">url_platform_subdomain_length</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:9%"></div></div><span class="feature-importance">0.0397</span></div>
<div class="feature-row"><span class="feature-rank">4</span><span class="feature-name">url_domain_dots</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:7%"></div></div><span class="feature-importance">0.0315</span></div>
<div class="feature-row"><span class="feature-rank">5</span><span class="feature-name">html_has_fromcharcode</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:7%"></div></div><span class="feature-importance">0.0296</span></div>
<div class="feature-row"><span class="feature-rank">6</span><span class="feature-name">url_num_domain_parts</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:6%"></div></div><span class="feature-importance">0.0269</span></div>
<div class="feature-row"><span class="feature-rank">7</span><span class="feature-name">html_has_meta_refresh</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:3%"></div></div><span class="feature-importance">0.0148</span></div>
<div class="feature-row"><span class="feature-rank">8</span><span class="feature-name">url_is_http</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:3%"></div></div><span class="feature-importance">0.0126</span></div>
<div class="feature-row"><span class="feature-rank">9</span><span class="feature-name">url_encoding_diff</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:3%"></div></div><span class="feature-importance">0.0124</span></div>
<div class="feature-row"><span class="feature-rank">10</span><span class="feature-name">url_path_digits</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:3%"></div></div><span class="feature-importance">0.0116</span></div>
<div class="feature-row"><span class="feature-rank">11</span><span class="feature-name">html_text_length</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:2%"></div></div><span class="feature-importance">0.0107</span></div>
<div class="feature-row"><span class="feature-rank">12</span><span class="feature-name">url_path_slashes</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:2%"></div></div><span class="feature-importance">0.0105</span></div>
<div class="feature-row"><span class="feature-rank">13</span><span class="feature-name">url_multiple_brands_in_url</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:2%"></div></div><span class="feature-importance">0.0103</span></div>
<div class="feature-row"><span class="feature-rank">14</span><span class="feature-name">url_brand_in_path</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:2%"></div></div><span class="feature-importance">0.0102</span></div>
<div class="feature-row"><span class="feature-rank">15</span><span class="feature-name">url_domain_hyphens</span><div class="importance-bar-bg"><div class="importance-bar-fill" style="width:2%"></div></div><span class="feature-importance">0.0095</span></div>
</div>
</div>
<div class="subsection">
<div class="section-title">Hyperparameters</div>
<div class="params-grid">
<div class="param-item"><span class="param-key">n_estimators</span><span class="param-value">726</span></div>
<div class="param-item"><span class="param-key">max_depth</span><span class="param-value">6</span></div>
<div class="param-item"><span class="param-key">learning_rate</span><span class="param-value">0.137</span></div>
<div class="param-item"><span class="param-key">subsample</span><span class="param-value">0.698</span></div>
</div>
</div>
</div>
</div>
<!-- CNN MODELS TAB -->
<div id="cnnModels" class="tab-content">
<div class="section-subtitle">
2 character-level CNN models that process raw text directly — no hand-crafted features needed.
Parallel Conv1D branches with kernel sizes (3, 5, 7) capture patterns at different scales.
</div>
<!-- CNN URL -->
<div class="model-detail">
<div class="model-detail-header">
<div class="model-detail-name">CNN URL (Char-level)</div>
<div class="model-detail-type">Deep Learning</div>
</div>
<div class="metrics-grid">
<div class="metric-card">
<div class="metric-value">98.38%</div>
<div class="metric-label">Accuracy</div>
</div>
<div class="metric-card">
<div class="metric-value">98.88%</div>
<div class="metric-label">Precision</div>
</div>
<div class="metric-card">
<div class="metric-value">97.86%</div>
<div class="metric-label">Recall</div>
</div>
<div class="metric-card">
<div class="metric-value">98.37%</div>
<div class="metric-label">F1-Score</div>
</div>
<div class="metric-card">
<div class="metric-value highlight">0.9976</div>
<div class="metric-label">ROC-AUC</div>
</div>
</div>
<div class="section-title">Confusion Matrix</div>
<div class="confusion-matrix">
<div class="cm-header"></div>
<div class="cm-header">Pred Legit</div>
<div class="cm-header">Pred Phish</div>
<div class="cm-label">Actual Legit</div>
<div class="cm-cell cm-tn">8,013</div>
<div class="cm-cell cm-fp">90</div>
<div class="cm-label">Actual Phish</div>
<div class="cm-cell cm-fn">173</div>
<div class="cm-cell cm-tp">7,930</div>
</div>
<div class="subsection">
<div class="section-title">Architecture</div>
<div class="params-grid">
<div class="param-item"><span class="param-key">Input</span><span class="param-value">Raw URL characters</span></div>
<div class="param-item"><span class="param-key">max_len</span><span class="param-value">800</span></div>
<div class="param-item"><span class="param-key">vocab_size</span><span class="param-value">87</span></div>
<div class="param-item"><span class="param-key">Conv1D kernels</span><span class="param-value">3, 5, 7</span></div>
<div class="param-item"><span class="param-key">Dataset</span><span class="param-value">108,034 URLs</span></div>
</div>
</div>
</div>
<!-- CNN HTML -->
<div class="model-detail">
<div class="model-detail-header">
<div class="model-detail-name">CNN HTML (Char-level)</div>
<div class="model-detail-type">Deep Learning</div>
</div>
<div class="metrics-grid">
<div class="metric-card">
<div class="metric-value">96.33%</div>
<div class="metric-label">Accuracy</div>
</div>
<div class="metric-card">
<div class="metric-value">98.18%</div>
<div class="metric-label">Precision</div>
</div>
<div class="metric-card">
<div class="metric-value">94.41%</div>
<div class="metric-label">Recall</div>
</div>
<div class="metric-card">
<div class="metric-value">96.26%</div>
<div class="metric-label">F1-Score</div>
</div>
<div class="metric-card">
<div class="metric-value highlight">0.9908</div>
<div class="metric-label">ROC-AUC</div>
</div>
</div>
<div class="section-title">Confusion Matrix</div>
<div class="confusion-matrix">
<div class="cm-header"></div>
<div class="cm-header">Pred Legit</div>
<div class="cm-header">Pred Phish</div>
<div class="cm-label">Actual Legit</div>
<div class="cm-cell cm-tn">5,943</div>
<div class="cm-cell cm-fp">106</div>
<div class="cm-label">Actual Phish</div>
<div class="cm-cell cm-fn">338</div>
<div class="cm-cell cm-tp">5,711</div>
</div>
<div class="subsection">
<div class="section-title">Architecture</div>
<div class="params-grid">
<div class="param-item"><span class="param-key">Input</span><span class="param-value">Raw HTML source</span></div>
<div class="param-item"><span class="param-key">max_len</span><span class="param-value">5,000</span></div>
<div class="param-item"><span class="param-key">vocab_size</span><span class="param-value">100</span></div>
<div class="param-item"><span class="param-key">Conv1D kernels</span><span class="param-value">3, 5, 7</span></div>
<div class="param-item"><span class="param-key">Dataset</span><span class="param-value">80,652 HTML pages</span></div>
</div>
</div>
</div>
</div>
<!-- COMPARISON TAB -->
<div id="overview" class="tab-content">
<div class="section-subtitle">
Side-by-side comparison of all 9 models across URL, HTML, Combined, and CNN categories.
</div>
<div class="section-title">All Models</div>
<div class="table-scroll">
<table class="comparison-table">
<thead>
<tr>
<th>Model</th>
<th>Category</th>
<th>Accuracy</th>
<th>Precision</th>
<th>Recall</th>
<th>F1-Score</th>
<th>ROC-AUC</th>
<th>Features</th>
</tr>
</thead>
<tbody>
<tr>
<td class="model-name-cell">Logistic Regression</td>
<td>URL</td>
<td>93.71%</td>
<td>95.40%</td>
<td>91.84%</td>
<td>93.59%</td>
<td>0.9789</td>
<td>121</td>
</tr>
<tr>
<td class="model-name-cell">Random Forest</td>
<td>URL</td>
<td>97.71%</td>
<td>99.06%</td>
<td>96.33%</td>
<td>97.68%</td>
<td>0.9958</td>
<td>121</td>
</tr>
<tr>
<td class="model-name-cell">XGBoost</td>
<td>URL</td>
<td>98.07%</td>
<td>99.12%</td>
<td>97.00%</td>
<td>98.05%</td>
<td>0.9963</td>
<td>121</td>
</tr>
<tr>
<td class="model-name-cell">Random Forest HTML</td>
<td>HTML</td>
<td>88.03%</td>
<td>87.49%</td>
<td>88.74%</td>
<td>88.11%</td>
<td>0.9561</td>
<td>100</td>
</tr>
<tr>
<td class="model-name-cell">XGBoost HTML</td>
<td>HTML</td>
<td>87.86%</td>
<td>86.45%</td>
<td>89.80%</td>
<td>88.09%</td>
<td>0.9557</td>
<td>100</td>
</tr>
<tr>
<td class="model-name-cell">RF Combined</td>
<td>Combined</td>
<td>98.60%</td>
<td>99.16%</td>
<td>98.02%</td>
<td>98.59%</td>
<td>0.9990</td>
<td>221</td>
</tr>
<tr>
<td class="model-name-cell">XGBoost Combined</td>
<td>Combined</td>
<td class="best">99.01%</td>
<td class="best">99.35%</td>
<td class="best">98.66%</td>
<td class="best">99.01%</td>
<td class="best">0.9991</td>
<td>221</td>
</tr>
<tr>
<td class="model-name-cell">CNN URL</td>
<td>CNN</td>
<td>98.38%</td>
<td>98.88%</td>
<td>97.86%</td>
<td>98.37%</td>
<td>0.9976</td>
<td>chars</td>
</tr>
<tr>
<td class="model-name-cell">CNN HTML</td>
<td>CNN</td>
<td>96.33%</td>
<td>98.18%</td>
<td>94.41%</td>
<td>96.26%</td>
<td>0.9908</td>
<td>chars</td>
</tr>
</tbody>
</table>
</div>
<div class="section-title">Key Insights</div>
<div class="insights-grid">
<div class="insight-card insight-safe">
<div class="insight-label">Best Overall</div>
<div class="insight-title">XGBoost Combined</div>
<div class="insight-desc">99.01% accuracy, 99.35% precision — best performance by combining 121 URL + 100 HTML features.</div>
</div>
<div class="insight-card insight-safe">
<div class="insight-label">Ensemble Strength</div>
<div class="insight-title">9-Model Consensus</div>
<div class="insight-desc">Combining 3 URL + 2 HTML + 2 Combined + 2 CNN models via majority vote maximizes reliability.</div>
</div>
<div class="insight-card insight-accent">
<div class="insight-label">Top Signal</div>
<div class="insight-title">html_num_links</div>
<div class="insight-desc">Number of links in HTML dominates XGBoost Combined at 44.2% importance — the single strongest feature.</div>
</div>
</div>
</div>
</section>
<footer>
<div class="footer-text">Machine Learning Phishing Detection</div>
</footer>
</div>
<script src="/static/script.js?v=4"></script>
</body>
</html>
|