Commit af0790c9dddbc170d5a251ccdd6283f1350313a2

Authored by Miguel Tuñón
1 parent 79d24d2f

Only conclusions left

Showing 89 changed files with 1002 additions and 653 deletions
ISO27001effectiveness/NAMESPACE
1 1 # Generated by roxygen2: do not edit by hand
2 2  
  3 +export(GetAttackTypeEvolution)
  4 +export(GetAttackTypePie)
  5 +export(GetAttackTypeSigleEvolution)
  6 +export(GetAttacksEvolution)
  7 +export(GetAttacksMonthEvolution)
  8 +export(GetCertsEvolution)
  9 +export(GetContinentAttackCol)
  10 +export(GetContinentAttacksEvolution)
  11 +export(GetContinentAttacksSigleEvolution)
  12 +export(GetContinentCertsEvolution)
  13 +export(GetContinentCertsSingleEvolution)
  14 +export(GetContinentPie)
  15 +export(GetCountriesAttacksSingleEvolution)
  16 +export(GetCountriesCertsSingleEvolution)
3 17 export(GetDefaultAttacksData)
4 18 export(GetISOSurveyCertsPerCountry)
5 19 export(GetISOSurveyCertsPerSector)
6 20 export(GetISOSurveySitesPerCountry)
7   -export(GetReportGraphs)
8 21 export(ParseHMExcel)
9 22 export(ParseHMFolder)
10 23 export(ProccesISOSurveyByCountryRaw)
... ...
ISO27001effectiveness/R/ReportGraphs.R
... ... @@ -301,6 +301,17 @@ GetAttackTypeSigleEvolution <- function(Attacks,
301 301 #-------------------------Geolocal evolution---------
302 302 #----------------------------------------------------------------
303 303  
  304 +#' Return pie graph to show % of attacks and certifications for each continent
  305 +#'
  306 +#' @param Attacks data.frame with procesed source data of attacks
  307 +#' @param Attacks.label.vjust Vector of vertical just to each portion label
  308 +#' @param Attacks.label.hjust Vector of horizontal just to each portion label
  309 +#' @param Cert_PerCountry data.frame with procesed source data of certifications
  310 +#' @param Certs.label.vjust Vector of vertical just to each portion label
  311 +#' @param Certs.label.hjust Vector of horizontal just to each portion label
  312 +#'
  313 +#' @return list(AttacksGraph, CertsGraph, ContinentListToStudy)
  314 +#' @export
304 315 GetContinentPie <- function (Attacks,
305 316 Attacks.label.vjust = 0,
306 317 Attacks.label.hjust = 0,
... ... @@ -362,19 +373,33 @@ GetContinentPie &lt;- function (Attacks,
362 373 axis.title.y=element_blank()) +
363 374 ggtitle("ISO 27001")
364 375  
365   - list(graph1, unique(attack.pie[attack.pie$perc > 5,]$Continent),
366   - graph2, unique(cert.pie[cert.pie$perc > 5,]$Continent))
  376 + #Return graphs and union of continent to study
  377 + list(graph1, graph2,
  378 + union(unique(attack.pie[attack.pie$perc > 5,]$Continent),
  379 + unique(cert.pie[cert.pie$perc > 5,]$Continent)))
367 380 }
368 381  
369   -GetContinentAttacksEvolution <- function(Attacks){
370 382  
371   - attacks.evol <- mutate(Attacks, Year = format(Attacks$Date, "%Y")) %>% group_by(Continent, Year)
  383 +#' Return graph to show the evolution of a attack types list
  384 +#'
  385 +#' @param Attacks data.frame with procesed source data
  386 +#' @param ContinentList List with continents to show
  387 +#'
  388 +#' @return list(graph, data.frame with ContinentList attacks data by year)
  389 +#' @export
  390 +GetContinentAttacksEvolution <- function(Attacks,
  391 + ContinentList){
  392 +
  393 + #Extract year from date
  394 + attacks.evol <- mutate(Attacks, Year = format(Attacks$Date, "%Y")) %>%
  395 + group_by(Continent, Year) #Grouping atacks by continent and year
  396 + #Counting attacks for each continent and year
372 397 attacks.evol <- as.data.frame(table(attacks.evol$Continent, attacks.evol$Year))
373 398 colnames(attacks.evol) <- c("Continent", "Year","Attacks")
  399 + #Filtering by the ContinentList especified
  400 + attacks.evol <- attacks.evol[attacks.evol$Continent %in% ContinentList,]
374 401  
375   - attacks.evol <- attacks.evol[attacks.evol$Continent != "Oceania",]
376   - attacks.evol <- attacks.evol[attacks.evol$Continent != "Africa",]
377   -
  402 + #Graph
378 403 graph1 <- ggplot2::qplot(main = "Cyberattacks evolution",
379 404 x = attacks.evol$Year,
380 405 y = attacks.evol$Attacks,
... ... @@ -384,26 +409,34 @@ GetContinentAttacksEvolution &lt;- function(Attacks){
384 409 data = attacks.evol,
385 410 geom = "point",
386 411 color = Continent) +
387   - geom_line() +
388   - theme(plot.title = element_text(hjust = 0.5))
  412 + geom_line() +
  413 + theme(plot.title = element_text(hjust = 0.5))
389 414  
390   - graph1
  415 + #Return graph and data to represent 1b1
  416 + list(graph1, attacks.evol)
391 417  
392 418 }
393 419  
394   -GetContinentCertsEvolution <- function(Cert_PerCountry){
395   -
396   - certs.evol <- gather(Cert_PerCountry, "Year", "Certs", 2:6) %>% group_by(Continent, Year)
397   -
398   - certs.evol <- summarise(certs.evol,
399   - #Continent = Continent,
400   - #Year = Year,
401   - Certs = sum(Certs))
  420 +#' Return graph to show the evolution of certifications in a continent list
  421 +#'
  422 +#' @param Cert_PerCountry data.frame with procesed source data
  423 +#' @param ContinentList List with continents to show
  424 +#'
  425 +#' @return list(graph, data.frame with ContinentList certifications data by year)
  426 +#' @export
  427 +GetContinentCertsEvolution <- function(Cert_PerCountry,
  428 + ContinentList){
  429 + #Collapsing year columns to 1 column with year value
  430 + certs.evol <- gather(Cert_PerCountry, "Year", "Certs", 2:6) %>%
  431 + group_by(Continent, Year) #Grouping by continent and year
  432 + #Sum of certifications for echa continent and year
  433 + certs.evol <- summarise(certs.evol, Certs = sum(Certs))
  434 + #Removing the X from year values
402 435 certs.evol$Year <- substr(certs.evol$Year,2,5)
  436 + #Filtering for the specified continents
  437 + certs.evol <- certs.evol[certs.evol$Continent %in% ContinentList,]
403 438  
404   - certs.evol <- certs.evol[certs.evol$Continent != "Oceania",]
405   - certs.evol <- certs.evol[certs.evol$Continent != "Africa",]
406   -
  439 + #graph
407 440 graph1 <- ggplot2::qplot(main = "ISO 27001 evolution",
408 441 x = certs.evol$Year,
409 442 y = certs.evol$Certs,
... ... @@ -413,268 +446,261 @@ GetContinentCertsEvolution &lt;- function(Cert_PerCountry){
413 446 data = certs.evol,
414 447 geom = "point",
415 448 color = Continent) +
416   - geom_line() +
417   - theme(plot.title = element_text(hjust = 0.5))
  449 + geom_line() +
  450 + theme(plot.title = element_text(hjust = 0.5))
418 451  
419   - graph1
  452 + #Return the graph and data to represent continents 1b1
  453 + list(graph1, certs.evol)
420 454  
421 455 }
  456 +#' Return graph to show the evolution of attacks in a single continent and his smooth
  457 +#'
  458 +#' @param Attacks data.frame with procesed source data
  459 +#' @param Continent Continent to represent
  460 +#' @param smooth.x Horizontal position to smooth slope label
  461 +#' @param smooth.y Vertical position to smooth slope label
  462 +#' @param smooth.label.hjust Horizontal just to smooth slope label
  463 +#'
  464 +#' @return list(graph, slope)
  465 +#' @export
  466 +GetContinentAttacksSigleEvolution <- function(Attacks,
  467 + Continent,
  468 + smooth.x = 0,
  469 + smooth.y = 0,
  470 + smooth.label.hjust = 0){
422 471  
423   -GetContinentAttacksTopEvolution <- function(Attacks){
424   - attacks.evol <- mutate(Attacks, Year = format(Attacks$Date, "%Y")) %>% group_by(Year, Continent)
425   - attacks.evol <- as.data.frame(table(attacks.evol$Year, attacks.evol$Continent))
426   - attacks.evol <- setNames(attacks.evol, c("Year", "Continent", "Count"))
427   - attacks.evol <- attacks.evol[attacks.evol$Continent != "",]
428   -
429   - attacks.evol <- spread(attacks.evol, "Continent", "Count")
  472 + #Filtering for the continent specified
  473 + attacks.evol <- Attacks[Attacks$Continent == Continent,]
430 474  
  475 + #Calc slope
  476 + slope1 <- lm(formula = Attacks ~ as.numeric(Year), data = attacks.evol)$coef[[2]]
  477 + #Graph
431 478 graph1 <- ggplot(data = attacks.evol,
432   - aes(x = Year, y = Americas, group = 1)) +
433   -
434   - geom_line() +
435   - geom_point() +
436   -
437   - theme(plot.title = element_text(hjust = 0.5)) +
438   - ggtitle("Americas") +
439   - xlab("Years") + ylab("Attacks")+
440   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
441   -
442   - graph2 <- ggplot(data = attacks.evol,
443   - aes(x = Year, y = Asia, group = 1)) +
444   -
445   - geom_line() +
446   - geom_point() +
447   -
448   - theme(plot.title = element_text(hjust = 0.5)) +
449   - ggtitle("Asia") +
450   - xlab("Years") + ylab("Attacks")+
451   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
452   -
453   - graph3 <- ggplot(data = attacks.evol,
454   - aes(x = Year, y = Europe, group = 1)) +
455   -
456   - geom_line() +
457   - geom_point() +
458   -
459   - theme(plot.title = element_text(hjust = 0.5)) +
460   - ggtitle("Europe") +
461   - xlab("Years") + ylab("Attacks")+
462   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
463   -
464   -
465   - list(graph1, graph2, graph3)
  479 + aes(x = Year, y = Attacks, group = 1)) +
  480 + geom_line() +
  481 + geom_point() +
  482 + theme(plot.title = element_text(hjust = 0.5)) +
  483 + ggtitle(Continent) +
  484 + xlab("Years") + ylab("Attacks")+
  485 + geom_smooth(method = "lm",
  486 + se = FALSE) +
  487 + geom_label(aes(x = smooth.x, y = smooth.y,
  488 + label = paste("Slope =", signif(slope1, 5))),
  489 + color = "blue",
  490 + hjust = smooth.label.hjust)
466 491  
  492 + #Returning list(graph, slope)
  493 + list(graph1, slope1)
467 494 }
  495 +#' Return graph to show the evolution of certifications in a single continent and his smooth
  496 +#'
  497 +#' @param Certs_byContinent data.frame with procesed source data
  498 +#' @param Continent Continent to represent
  499 +#' @param smooth.x Horizontal position to smooth slope label
  500 +#' @param smooth.y Vertical position to smooth slope label
  501 +#' @param smooth.label.hjust Horizontal just to smooth slope label
  502 +#'
  503 +#' @return list(graph, slope)
  504 +#' @export
  505 +GetContinentCertsSingleEvolution <- function(Certs_byContinent,
  506 + Continent,
  507 + smooth.x = 0,
  508 + smooth.y = 0,
  509 + smooth.label.hjust = 0){
468 510  
469   -GetContinentCertsTopEvolution <- function(Attacks){
470   -
471   - certs.evol <- gather(Cert_PerCountry, "Year", "Certs", 2:6) %>% group_by(Continent, Year)
472   -
473   - certs.evol <- summarise(certs.evol,
474   - #Continent = Continent,
475   - #Year = Year,
476   - Certs = sum(Certs))
477   - certs.evol$Year <- substr(certs.evol$Year,2,5)
478   -
479   - certs.evol <- spread(certs.evol, "Continent", "Certs")
  511 + #Filtering for the continent specified
  512 + certs.evol <- Certs_byContinent[Certs_byContinent$Continent == Continent,]
480 513  
  514 + #Calc slope
  515 + slope1 <- lm(formula = Certs ~ as.numeric(Year), data = certs.evol)$coef[[2]]
  516 + #Graph
481 517 graph1 <- ggplot(data = certs.evol,
482   - aes(x = Year, y = Americas, group = 1)) +
483   -
484   - geom_line() +
485   - geom_point() +
486   -
487   - theme(plot.title = element_text(hjust = 0.5)) +
488   - ggtitle("Americas") +
489   - xlab("Years") + ylab("Certifications")+
490   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
491   -
492   - graph2 <- ggplot(data = certs.evol,
493   - aes(x = Year, y = Asia, group = 1)) +
494   -
495   - geom_line() +
496   - geom_point() +
497   -
498   - theme(plot.title = element_text(hjust = 0.5)) +
499   - ggtitle("Asia") +
500   - xlab("Years") + ylab("Certifications")+
501   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
502   -
503   - graph3 <- ggplot(data = certs.evol,
504   - aes(x = Year, y = Europe, group = 1)) +
505   -
506   - geom_line() +
507   - geom_point() +
508   -
509   - theme(plot.title = element_text(hjust = 0.5)) +
510   - ggtitle("Europe") +
511   - xlab("Years") + ylab("Certifications")+
512   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
513   -
514   -
515   - list(graph1, graph2, graph3)
  518 + aes(x = Year, y = Certs, group = 1)) +
  519 + geom_line() +
  520 + geom_point() +
  521 + theme(plot.title = element_text(hjust = 0.5)) +
  522 + ggtitle(Continent) +
  523 + xlab("Years") + ylab("Certifications")+
  524 + geom_smooth(method = "lm",
  525 + se = FALSE) +
  526 + geom_label(aes(x = smooth.x, y = smooth.y,
  527 + label = paste("Slope =", signif(slope1, 5))),
  528 + color = "blue",
  529 + hjust = smooth.label.hjust)
516 530  
  531 + #Returning list(graph, slope)
  532 + list(graph1, slope1)
517 533 }
518 534  
519   -GetCountriesCol <- function(Attacks, Cert_PerCountry){
520   -
521   - certs.evol <- gather(Cert_PerCountry, "Year", "Certs", 2:6) %>% group_by(Continent, country_short)
522   -
523   - certs.evol <- summarise(certs.evol,
524   - Certs = sum(Certs))
525   -
526   - graph1 <- ggplot2::ggplot(aes(x = reorder(country_short, Certs),
527   - y = Certs),
528   - data = certs.evol[certs.evol$Certs > (sum(certs.evol$Certs) * 0.02),]) +
529   - geom_col(aes(fill = Continent)) +
530   - theme(plot.title = element_text(hjust = 0.5)) +
531   - ggtitle("ISO 27001") +
532   - xlab("Country") + ylab("Certifications")
  535 +#' Return graph to show the attacks by country
  536 +#'
  537 +#' @param Attacks data.frame with procesed source data of attacks
  538 +#' @param Attacks.label.vjust Vector of vertical just to each portion label
  539 +#' @param Attacks.label.hjust Vector of horizontal just to each portion label
  540 +#' @param Cert_PerCountry data.frame with procesed source certifications
  541 +#' @param Certs.label.vjust Vector of vertical just to each portion label
  542 +#' @param Certs.label.hjust Vector of horizontal just to each portion label
  543 +#'
  544 +#' @return list(AttacksGraph, CertsGraph, CountryListToStudy)
  545 +GetCountriesCol <- function(Attacks,
  546 + Attacks.label.vjust = 0,
  547 + Attacks.label.hjust = 0,
  548 + Cert_PerCountry,
  549 + Certs.label.vjust = 0,
  550 + Certs.label.hjust = 0){
  551 +
  552 + #Group attacks by continent and country
  553 + attacks.col <- group_by(Attacks, Continent, Country)
  554 + #Count attacks for each country
  555 + attacks.col <- as.data.frame(table(attacks.col$Continent, attacks.col$Country))
  556 + attacks.col <- setNames(attacks.col, c("Continent", "Country", "Count"))
  557 + #Remove rows without Continent/Country specified
  558 + attacks.col <- attacks.col[attacks.col$Country != "",]
  559 + attacks.col <- attacks.col[attacks.col$Country != "AU",] #Abnormal behaivour, corrupt??
  560 + attacks.col <- attacks.col[attacks.col$Continent != "",]
  561 + #Only the countries with more than 2% of total attacks
  562 + attacks.col <- attacks.col[attacks.col$Count > (sum(attacks.col$Count) * 0.015), ]
  563 + #Sort by attacks
  564 + attacks.col <- arrange(attacks.col, desc(Count))
533 565  
534   - attacks.evol <- mutate(Attacks, Year = format(Attacks$Date, "%Y")) %>% group_by(Year, Continent, Country)
535   - attacks.evol <- as.data.frame(table(attacks.evol$Year, attacks.evol$Continent, attacks.evol$Country))
536   - attacks.evol <- setNames(attacks.evol, c("Year", "Continent", "Country", "Count"))
537   - attacks.evol <- attacks.evol[attacks.evol$Continent != "",]
538   - attacks.evol <- attacks.evol[attacks.evol$Country != "",]
539   - attacks.evol <- attacks.evol[attacks.evol$Country != "AU",]
  566 + #Attacks graph
  567 + graph1 <- ggplot2::ggplot(aes(x = reorder(Country, Count),
  568 + y = Count),
  569 + data = attacks.col) +
  570 + geom_col(aes(fill = Continent)) +
  571 + geom_text(aes(label = attacks.col$Count),
  572 + vjust = Attacks.label.vjust,
  573 + hjust = Attacks.label.hjust,
  574 + size = 3) +
  575 + theme(plot.title = element_text(hjust = 0.5)) +
  576 + ggtitle("Attacks") +
  577 + xlab("Country") + ylab("Attacks")
540 578  
541   - attacks.evol <- attacks.evol[attacks.evol$Count > (sum((attacks.evol[attacks.evol$Country != "US",])$Count) * 0.01),]
542   - attacks.evol <- arrange(attacks.evol, Count)
  579 + #Grouping certifications by continent
  580 + certs.col <- group_by(Cert_PerCountry, Continent, country_short)
  581 + #Counting certificates for each continent
  582 + certs.col <- dplyr::summarise(certs.col, Count = sum(X2011 + X2012 + X2013 + X2014 + X2015))
  583 + #Remove rows without Continent specified
  584 + certs.col <- certs.col[certs.col$Continent != "",]
  585 + certs.col <- certs.col[certs.col$country_short != "",]
  586 + #Only the countries with more than 2% of total certs
  587 + certs.col <- certs.col[certs.col$Count > (sum(certs.col$Count) * 0.02), ]
  588 + #Sort by certifications
  589 + certs.col <- arrange(certs.col, desc(Count))
543 590  
544 591  
545   - graph2 <- ggplot2::ggplot(aes(x = reorder(Country, Count),
  592 + #Certifications graph
  593 + graph2 <- ggplot2::ggplot(aes(x = reorder(country_short, Count),
546 594 y = Count),
547   - data = attacks.evol) +
548   - geom_col(aes(fill = Continent)) +
549   - theme(plot.title = element_text(hjust = 0.5)) +
550   - ggtitle("Attacks") +
551   - xlab("Country") + ylab("Attacks")
  595 + data = certs.col) +
  596 + geom_col(aes(fill = Continent)) +
  597 + geom_text(aes(label = certs.col$Count),
  598 + vjust = Certs.label.vjust,
  599 + hjust = Certs.label.hjust,
  600 + size = 3) +
  601 + theme(plot.title = element_text(hjust = 0.5)) +
  602 + ggtitle("ISO 27001") +
  603 + xlab("Country") + ylab("Certifications")
552 604  
553   - list(graph1, graph2)
  605 + #Return graphs and union of countries to study
  606 + list(graph1, graph2,
  607 + union(unique(head(attacks.col$Country, 3)),
  608 + unique(head(certs.col$country_short, 3))))
554 609  
555 610 }
556 611  
557   -GetCountriesAttacksTopEvolution <- function(Attacks){
558   - attacks.evol <- mutate(Attacks, Year = format(Attacks$Date, "%Y")) %>% group_by(Year, Country)
  612 +#' Return graph to show the evolution of attacks in a single continent and his smooth
  613 +#'
  614 +#' @param Attacks data.frame with procesed source data
  615 +#' @param Country Country to represent
  616 +#' @param smooth.x Horizontal position to smooth slope label
  617 +#' @param smooth.y Vertical position to smooth slope label
  618 +#' @param smooth.label.hjust Horizontal just to smooth slope label
  619 +#'
  620 +#' @return list(graph, slope)
  621 +#' @export
  622 +GetCountriesAttacksSingleEvolution <- function(Attacks,
  623 + Country,
  624 + smooth.x = 0,
  625 + smooth.y = 0,
  626 + smooth.label.hjust = 0){
  627 + #Extract year from date
  628 + attacks.evol <- mutate(Attacks, Year = format(Attacks$Date, "%Y")) %>%
  629 + group_by(Year, Country) #Group by country
  630 + #Count attacks for each year and country
559 631 attacks.evol <- as.data.frame(table(attacks.evol$Year, attacks.evol$Country))
560 632 attacks.evol <- setNames(attacks.evol, c("Year", "Country", "Count"))
561   - attacks.evol <- attacks.evol[attacks.evol$Country != "",]
562   -
563   - attacks.evol <- spread(attacks.evol, "Country", "Count")
  633 + #Filter by the specified country
  634 + attacks.evol <- attacks.evol[attacks.evol$Country == Country,]
564 635  
565   - slope1 <- lm(formula = US ~ as.numeric(Year), data = attacks.evol)
  636 + #Calc slope
  637 + slope1 <- lm(formula = Count ~ as.numeric(Year), data = attacks.evol)$coef[[2]]
  638 + #Graph
566 639 graph1 <- ggplot(data = attacks.evol,
567   - aes(x = Year, y = US, group = 1)) +
568   -
569   - geom_line() +
570   - geom_point() +
571   -
572   - theme(plot.title = element_text(hjust = 0.5)) +
573   - ggtitle("US") +
574   - xlab("Years") + ylab("Attacks")+
575   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..)) +
576   - geom_label(aes(x = "2014", y = 700, label = signif(slope1$coef[[2]])), color = "blue")
577   -
578   - graph2 <- ggplot(data = attacks.evol,
579   - aes(x = Year, y = GB, group = 1)) +
580   -
581   - geom_line() +
582   - geom_point() +
583   -
584   - theme(plot.title = element_text(hjust = 0.5)) +
585   - ggtitle("GB") +
586   - xlab("Years") + ylab("Attacks")+
587   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
588   -
589   - graph3 <- ggplot(data = attacks.evol,
590   - aes(x = Year, y = IN, group = 1)) +
591   -
592   - geom_line() +
593   - geom_point() +
594   -
595   - theme(plot.title = element_text(hjust = 0.5)) +
596   - ggtitle("IN") +
597   - xlab("Years") + ylab("Attacks")+
598   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
599   -
600   - graph4 <- ggplot(data = attacks.evol,
601   - aes(x = Year, y = JP, group = 1)) +
602   -
603   - geom_line() +
604   - geom_point() +
605   -
606   - theme(plot.title = element_text(hjust = 0.5)) +
607   - ggtitle("JP") +
608   - xlab("Years") + ylab("Attacks")+
609   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
610   -
  640 + aes(x = Year, y = Count, group = 1)) +
  641 + geom_line() +
  642 + geom_point() +
  643 + theme(plot.title = element_text(hjust = 0.5)) +
  644 + ggtitle(Country) +
  645 + xlab("Years") + ylab("Attacks")+
  646 + geom_smooth(method = "lm",
  647 + se = FALSE) +
  648 + geom_label(aes(x = smooth.x, y = smooth.y,
  649 + label = paste("Slope =", signif(slope1, 5))),
  650 + color = "blue",
  651 + hjust = smooth.label.hjust)
611 652  
612   - list(graph1, graph2, graph3, graph4)
  653 + #Return list(graph, slope)
  654 + list(graph1, slope1)
613 655  
614 656 }
615 657  
616   -GetCountriesCertsTopEvolution <- function(Attacks){
617   -
618   - certs.evol <- gather(Cert_PerCountry, "Year", "Certs", 2:6) %>% group_by(country_short, Year)
619   -
620   - certs.evol <- rbind(certs.evol[certs.evol$country_short == "US",],
621   - certs.evol[certs.evol$country_short == "GB",],
622   - certs.evol[certs.evol$country_short == "IN",],
623   - certs.evol[certs.evol$country_short == "JP",])
624   -
  658 +#' Return graph to show the evolution of certifications in a single continent and his smooth
  659 +#'
  660 +#' @param Attacks data.frame with procesed source data
  661 +#' @param Country Country to represent
  662 +#' @param smooth.x Horizontal position to smooth slope label
  663 +#' @param smooth.y Vertical position to smooth slope label
  664 +#' @param smooth.label.hjust Horizontal just to smooth slope label
  665 +#'
  666 +#' @return list(graph, slope)
  667 +#' @export
  668 +GetCountriesCertsSingleEvolution <- function(Cert_PerCountry,
  669 + Country,
  670 + smooth.x = 0,
  671 + smooth.y = 0,
  672 + smooth.label.hjust = 0){
  673 +
  674 + #Collapsing year columns to only one with the year value
  675 + certs.evol <- gather(Cert_PerCountry, "Year", "Certs", 2:6) %>%
  676 + group_by(country_short, Year) #Group by country and year
  677 + #sum certificates for each country and year
625 678 certs.evol <- summarise(certs.evol,
626 679 Certs = sum(Certs))
  680 + #Removing the X of the year values
627 681 certs.evol$Year <- substr(certs.evol$Year,2,5)
  682 + #Filter by the specified Country
  683 + certs.evol <- certs.evol[certs.evol$country_short == Country,]
628 684  
629   - certs.evol <- spread(certs.evol, "country_short", "Certs")
630   -
  685 + #Calc slope
  686 + slope1 <- lm(formula = Certs ~ as.numeric(Year), data = certs.evol)$coef[[2]]
  687 + #Graph
631 688 graph1 <- ggplot(data = certs.evol,
632   - aes(x = Year, y = US, group = 1)) +
633   -
  689 + aes(x = Year, y = Certs, group = 1)) +
634 690 geom_line() +
635 691 geom_point() +
636   -
637 692 theme(plot.title = element_text(hjust = 0.5)) +
638   - ggtitle("US") +
  693 + ggtitle(Country) +
639 694 xlab("Years") + ylab("Certifications")+
640   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
641   -
642   - graph2 <- ggplot(data = certs.evol,
643   - aes(x = Year, y = GB, group = 1)) +
644   -
645   - geom_line() +
646   - geom_point() +
647   -
648   - theme(plot.title = element_text(hjust = 0.5)) +
649   - ggtitle("GB") +
650   - xlab("Years") + ylab("Certifications")+
651   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
652   -
653   - graph3 <- ggplot(data = certs.evol,
654   - aes(x = Year, y = IN, group = 1)) +
655   -
656   - geom_line() +
657   - geom_point() +
658   -
659   - theme(plot.title = element_text(hjust = 0.5)) +
660   - ggtitle("IN") +
661   - xlab("Years") + ylab("Certifications")+
662   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
663   -
664   -
665   - graph4 <- ggplot(data = certs.evol,
666   - aes(x = Year, y = JP, group = 1)) +
667   -
668   - geom_line() +
669   - geom_point() +
670   -
671   - theme(plot.title = element_text(hjust = 0.5)) +
672   - ggtitle("JP") +
673   - xlab("Years") + ylab("Certifications")+
674   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..))
675   -
676   -
677   - list(graph1, graph2, graph3, graph4)
  695 + geom_smooth(method = "lm",
  696 + se = FALSE) +
  697 + geom_label(aes(x = smooth.x, y = smooth.y,
  698 + label = paste("Slope =", signif(slope1, 5))),
  699 + color = "blue",
  700 + hjust = smooth.label.hjust)
  701 +
  702 + #Return graph and slope
  703 + list(graph1, slope1)
678 704  
679 705 }
680 706  
... ... @@ -683,300 +709,134 @@ GetCountriesCertsTopEvolution &lt;- function(Attacks){
683 709 #-------------------------gelocal/type---------------------------
684 710 #----------------------------------------------------------------
685 711  
686   -GetContinentAttackPie <- function (Attacks){
  712 +#' Return graph representing % of attacks type by continent
  713 +#'
  714 +#' @param Attacks data.frame with procesed source data
  715 +#' @param Country Country to represent on the left side
  716 +#' @param Country2 Country to represent on the right side
  717 +#'
  718 +#' @return Graph
  719 +#' @export
  720 +GetContinentAttackCol <- function (Attacks,
  721 + Country,
  722 + Country2){
687 723  
  724 + #Grouping attacks by attack type and country
688 725 attack.pie <- group_by(Attacks, Attack.standar, Country)
  726 + #Counting attacks for each attack type and country
689 727 attack.pie <- as.data.frame(table(attack.pie$Country, attack.pie$Attack.standar))
690 728 attack.pie <- setNames(attack.pie, c("Country", "Attack", "Count"))
691   -
  729 + #Remove rows withouth a country specified
692 730 attack.pie <- attack.pie[attack.pie$Country != "",]
  731 + #Remove rows withouth attack type specified
693 732 attack.pie <- attack.pie[attack.pie$Attack != "",]
694   -
  733 + #grouping non 'important' attack types in 'Otros'
695 734 attack.pie$Attack <- as.character(attack.pie$Attack)
696 735 attack.pie[attack.pie$Attack != "DDoS" &
697 736 attack.pie$Attack != "Defacement" &
698 737 attack.pie$Attack != "Injection",]$Attack <- "Otros"
699 738 attack.pie$Attack <- as.factor(attack.pie$Attack)
700   -
  739 + #Grouping atatcks by attacky type and country
701 740 attack.pie <- group_by(attack.pie, Attack, Country)
  741 + #sum attacks for each attack type and country
702 742 attack.pie <- summarise(attack.pie, Count = sum(Count))
703 743  
704   - attack.pie.US <- attack.pie[attack.pie$Country == "US",]
705   - attack.pie.JP <- attack.pie[attack.pie$Country == "JP",]
  744 + #Filtering by the desired countries
  745 + attack.pie.C1 <- attack.pie[attack.pie$Country == Country,]
  746 + attack.pie.C2 <- attack.pie[attack.pie$Country == Country2,]
706 747  
  748 + #Calc % of each attack type
  749 + attack.pie.C1$perc <- round(100 * attack.pie.C1$Count / sum(attack.pie.C1$Count), 2)
  750 + attack.pie.C2$perc <- round(100 * attack.pie.C2$Count / sum(attack.pie.C2$Count), 2)
  751 +
  752 + #Calc max % to draw gray background
  753 + attack.pie.max <- data.frame(Attack = attack.pie.C1$Attack)
  754 + percs <- c()
  755 + for (AT in attack.pie.max$Attack){
  756 + percs <- c(percs, max(attack.pie.C1[attack.pie.C1$Attack == AT,]$perc,
  757 + attack.pie.C2[attack.pie.C2$Attack == AT,]$perc))
  758 + }
  759 + attack.pie.max$perc <- percs
  760 +
  761 + graph1 <- ggplot() +
  762 + geom_col(aes(x=Attack,
  763 + y=perc),
  764 + width = 1,
  765 + data = attack.pie.max,
  766 + color = "Black",
  767 + fill = "Grey") +
  768 + geom_col(aes(x=1.25:4.25,
  769 + y=perc,
  770 + fill = Country),
  771 + width = 0.4,
  772 + data = attack.pie.C1) +
  773 + geom_text(aes(x=1.25:4.25,
  774 + y=perc + 1,
  775 + label = paste(perc, "%")),
  776 + data = attack.pie.C1) +
  777 + geom_col(aes(x=0.75:3.75,
  778 + y=perc,
  779 + fill = Country2),
  780 + width = 0.4,
  781 + data = attack.pie.C2) +
  782 + geom_text(aes(x=0.75:3.75,
  783 + y=perc + 1,
  784 + label = paste(perc, "%")),
  785 + data = attack.pie.C2) +
  786 + scale_x_discrete("Attack", attack.pie.C1$Attack) +
  787 + ylab("% of total attacks") +
  788 + theme(plot.title = element_text(hjust = 0.5)) +
  789 + ggtitle(paste(Country, Country2, sep = " - "))
707 790  
708   - graph1 <- ggplot(data=attack.pie.US,
709   - aes(x=factor(1),
710   - y=Count,
711   - fill=Attack)) +
712   - geom_col(width = 1, color='black') +
713   - geom_label(aes(label=paste(round(100 * attack.pie.US$Count / sum(attack.pie.US$Count), 2), "%")),
714   - vjust=c(0),
715   - hjust=c(0)) +
716   - coord_polar(theta="y") +
717   - scale_x_discrete(labels = c("")) +
718   - scale_y_discrete(labels = c("")) +
719   - theme(plot.title = element_text(hjust = 0.5),
720   - axis.title.x=element_blank(),
721   - axis.title.y=element_blank()) +
722   - ggtitle("US")
723   -
724   - graph2 <- ggplot(data=attack.pie.JP,
725   - aes(x=factor(1),
726   - y=Count,
727   - fill=Attack)) +
728   - geom_col(width = 1, color='black') +
729   - geom_label(aes(label=paste(round(100 * attack.pie.JP$Count / sum(attack.pie.JP$Count), 2), "%")),
730   - vjust=c(0),
731   - hjust=c(0)) +
732   - coord_polar(theta="y") +
733   - scale_x_discrete(labels = c("")) +
734   - scale_y_discrete(labels = c("")) +
735   - theme(plot.title = element_text(hjust = 0.5),
736   - axis.title.x=element_blank(),
737   - axis.title.y=element_blank()) +
738   - ggtitle("JP")
739   -
740   - list(graph1, graph2)
  791 + graph1
741 792 }
742 793  
743   -GetContinentAttackEvolution <- function(Attacks){
744 794  
  795 +#' Return graph representing the evolution of attack types in a country
  796 +#'
  797 +#' @param Attacks data.frame with procesed source data
  798 +#' @param Country Country to represent
  799 +#'
  800 +#' @return list(Graph, data to represent attacktypes of the country 1b1)
  801 +GetContinentAttackEvolution <- function(Attacks, Country){
  802 +
  803 + #Extract Year from date
745 804 attack.evol <-mutate(Attacks, Year = format(Attacks$Date, "%Y")) %>%
746   - group_by(Attack.standar, Country, Year)
  805 + group_by(Attack.standar, Country, Year) #Group by attack type, country and year
  806 + #Counting the attacks for each attack type, country and year
747 807 attack.evol <- as.data.frame(table(attack.evol$Country, attack.evol$Attack.standar, attack.evol$Year))
748 808 attack.evol <- setNames(attack.evol, c("Country", "Attack", "Year", "Count"))
749   -
  809 + #Removing rows without country or attack type
750 810 attack.evol <- attack.evol[attack.evol$Country != "",]
751 811 attack.evol <- attack.evol[attack.evol$Attack != "",]
752   -
  812 + #grouping non 'important' attack types in 'Otros'
753 813 attack.evol$Attack <- as.character(attack.evol$Attack)
754 814 attack.evol[attack.evol$Attack != "DDoS" &
755 815 attack.evol$Attack != "Defacement" &
756 816 attack.evol$Attack != "Injection",]$Attack <- "Otros"
757 817 attack.evol$Attack <- as.factor(attack.evol$Attack)
758   -
  818 + #Counting attacks of new tyoe
759 819 attack.evol <- group_by(attack.evol, Attack, Country, Year)
760 820 attack.evol <- summarise(attack.evol, Count = sum(Count))
  821 + #Filtering by specified country
  822 + attack.evol <- attack.evol[attack.evol$Country == Country,]
761 823  
762   - attack.evol.US <- attack.evol[attack.evol$Country == "US",]
763   - attack.evol.JP <- attack.evol[attack.evol$Country == "JP",]
764   -
765   - attack.evol.US.Otros <- attack.evol.US[attack.evol.US$Attack == "Otros",]
766   - attack.evol.JP.Otros <- attack.evol.JP[attack.evol.JP$Attack == "Otros",]
767   -
768   -
769   - graph1 <- ggplot2::qplot(main = "US",
770   - x = attack.evol.US$Year,
771   - y = attack.evol.US$Count,
772   - group = Attack,
773   - xlab = "Years",
774   - ylab = "Certifications",
775   - data = attack.evol.US,
776   - geom = "point",
777   - color = Attack) +
778   - geom_line() +
779   - theme(plot.title = element_text(hjust = 0.5))
780   -
781   - graph2 <- ggplot2::qplot(main = "JP",
782   - x = attack.evol.JP$Year,
783   - y = attack.evol.JP$Count,
784   - group = Attack,
  824 + #Graph
  825 + graph1 <- ggplot2::qplot(main = Country,
  826 + x = attack.evol$Year,
  827 + y = attack.evol$Count,
  828 + group = attack.evol$Attack,
785 829 xlab = "Years",
786   - ylab = "Certifications",
787   - data = attack.evol.JP,
  830 + ylab = "Attacks",
  831 + data = attack.evol,
788 832 geom = "point",
789 833 color = Attack) +
790 834 geom_line() +
791 835 theme(plot.title = element_text(hjust = 0.5))
792 836  
793   - graph3 <- ggplot2::qplot(main = "US",
794   - x = attack.evol.US.Otros$Year,
795   - y = attack.evol.US.Otros$Count,
796   - group = 1,
797   - xlab = "Years",
798   - ylab = "Certifications",
799   - data = attack.evol.US.Otros,
800   - geom = "point") +
801   - geom_line() +
802   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..)) +
803   - theme(plot.title = element_text(hjust = 0.5))
804   -
805   - graph4 <- ggplot2::qplot(main = "JP",
806   - x = attack.evol.JP.Otros$Year,
807   - y = attack.evol.JP.Otros$Count,
808   - group = 1,
809   - xlab = "Years",
810   - ylab = "Certifications",
811   - data = attack.evol.JP.Otros,
812   - geom = "point") +
813   - geom_line() +
814   - stat_smooth(method = "lm", se = FALSE, aes(outfit=fit<<-..y..)) +
815   - theme(plot.title = element_text(hjust = 0.5))
816   -
817   - list(graph1, graph2, graph3, graph4)
  837 + #list with graph and processed data to represent 1b1
  838 + list(graph1, attack.evol)
818 839  
819 840 }
820 841  
821   -#----------------------------------------------------------------
822   -#----------------------------------------------------------------
823   -#----------------------------------------------------------------
824   -
825 842  
826   -GetBaseCertsGraph <- function(Cert_PerCountry, year){
827   - graph1 <- ggplot2::qplot(main = "Countries with above average number of companies certified with 27001 (2012)",
828   - x = reorder(country_short,X2012),
829   - y = X2012,
830   - xlab = "Country",
831   - ylab = "Number of certifications",
832   - data = Cert_PerCountry[Cert_PerCountry$X2012 > mean(Cert_PerCountry$X2012),],
833   - geom = "col",
834   - fill = Continent)
835   - graph1
836   -}
837   -
838   -
839   -
840   -
841   -#' Return every graph used in the report file
842   -#'
843   -#' @param Cert_PerCountry data.frame with the processed data of ISO 27001 certifications
844   -#' @param Attacks data.frame with the processed data of cyberattacks
845   -#'
846   -#' @return data.frame
847   -#' @export
848   -GetReportGraphs <- function(Cert_PerCountry,Attacks) {
849   - #2012
850   - graph1 <- ggplot2::qplot(main = "Countries with above average number of companies certified with 27001 (2012)",
851   - x = reorder(country_short,X2012),
852   - y = X2012,
853   - xlab = "Country",
854   - ylab = "Number of certifications",
855   - data = Cert_PerCountry[Cert_PerCountry$X2012 > mean(Cert_PerCountry$X2012),],
856   - geom = "col",
857   - fill = Continent)
858   -
859   - attacks2k12 <- Attacks[Attacks$Date < "2013-01-01" & Attacks$Date >= "2012-01-01",]
860   - frameAttacks2k12 <- as.data.frame(table(attacks2k12$Country))
861   - colnames(frameAttacks2k12) <- c("Country","Attacks")
862   - graph2 <- ggplot2::qplot(main = "Countries with above average number of cyberattacks (2012)",
863   - x = reorder(Country,Attacks),
864   - y = Attacks,
865   - xlab = "Country",
866   - ylab = "Number of attacks",
867   - data = frameAttacks2k12[frameAttacks2k12$Attacks > mean(frameAttacks2k12$Attacks),],
868   - geom = "col",
869   - fill = Continent)
870   -
871   - Attacks2012ByMonth <- mutate(attacks2k12, month = format(attacks2k12$Date, "%m")) %>% group_by(month)
872   - Attack2012FreqByMonth <- as.data.frame(table(Attacks2012ByMonth$month))
873   - colnames(Attack2012FreqByMonth) <- c("Month", "Attacks")
874   - graph3 <- ggplot2::qplot(x = as.numeric(Month),
875   - y = Attacks,
876   - main = "Global cyberattack progression by month (2012)",
877   - data = Attack2012FreqByMonth,
878   - geom = c("point", "smooth"),
879   - xlim = c(1,12),
880   - xlab = "Month") + ggplot2::scale_x_continuous(breaks = 1:12)
881   -
882   - #2013
883   - graph4 <- ggplot2::qplot(main = "Countries with above average number of companies certified with 27001 (2013)",
884   - x = reorder(country_short,X2013),
885   - y = X2013,
886   - xlab = "Country",
887   - ylab = "Number of certifications",
888   - data = Cert_PerCountry[Cert_PerCountry$X2013 > mean(Cert_PerCountry$X2013),]
889   - , geom = "col",
890   - fill = Continent)
891   - attacks2k13 <- Attacks[Attacks$Date < "2014-01-01" & Attacks$Date >= "2013-01-01",]
892   - frameAttacks2k13 <- as.data.frame(table(attacks2k13$Country))
893   - colnames(frameAttacks2k13) <- c("Country","Attacks")
894   - graph5 <- ggplot2::qplot(main = "Countries with above average number of cyberattacks (2013)",
895   - x = reorder(Country,Attacks),
896   - y = Attacks,
897   - xlab = "Country",
898   - ylab = "Number of attacks",
899   - data = frameAttacks2k13[frameAttacks2k13$Attacks > mean(frameAttacks2k13$Attacks),]
900   - , geom = "col",
901   - fill = Continent)
902   -
903   - Attacks2013ByMonth <- mutate(attacks2k13, month = format(attacks2k13$Date, "%m")) %>% group_by(month)
904   - Attack2013FreqByMonth <- as.data.frame(table(Attacks2013ByMonth$month))
905   - colnames(Attack2013FreqByMonth) <- c("Month", "Attacks")
906   - graph6 <- ggplot2::qplot(x = as.numeric(Month),
907   - y = Attacks,
908   - main = "Global cyberattack progression by month (2013)",
909   - data = Attack2013FreqByMonth,
910   - geom = c("point", "smooth"),
911   - xlim = c(1,12),
912   - xlab = "Month") + ggplot2::scale_x_continuous(breaks = 1:12)
913   -
914   - #2014
915   - graph7 <- ggplot2::qplot(main = "Countries with above average number of companies certified with 27001 (2014)",
916   - x = reorder(country_short,X2014),
917   - y = X2014,
918   - xlab = "Country",
919   - ylab = "Number of certifications",
920   - data = Cert_PerCountry[Cert_PerCountry$X2014 > mean(Cert_PerCountry$X2014),]
921   - , geom = "col",
922   - fill = Continent)
923   - attacks2k14 <- Attacks[Attacks$Date < "2015-01-01" & Attacks$Date >= "2014-01-01",]
924   - frameAttacks2k14 <- as.data.frame(table(attacks2k14$Country))
925   - colnames(frameAttacks2k14) <- c("Country","Attacks")
926   - graph8 <- ggplot2::qplot(main = "Countries with above average number of cyberattacks (2014)",
927   - x = reorder(Country,Attacks),
928   - y = Attacks,
929   - xlab = "Country",
930   - ylab = "Number of attacks",
931   - data = frameAttacks2k14[frameAttacks2k14$Attacks > mean(frameAttacks2k14$Attacks),]
932   - , geom = "col",
933   - fill = Continent)
934   -
935   - Attacks2014ByMonth <- mutate(attacks2k14, month = format(attacks2k14$Date, "%m")) %>% group_by(month)
936   - Attack2014FreqByMonth <- as.data.frame(table(Attacks2014ByMonth$month))
937   - colnames(Attack2014FreqByMonth) <- c("Month", "Attacks")
938   - graph9 <- ggplot2::qplot(x = as.numeric(Month),
939   - y = Attacks,
940   - main = "Global cyberattack progression by month (2014)",
941   - data = Attack2014FreqByMonth,
942   - geom = c("point", "smooth"),
943   - xlim = c(1,12),
944   - xlab = "Month") + ggplot2::scale_x_continuous(breaks = 1:12)
945   -
946   - #2015
947   - graph10 <- ggplot2::qplot(main = "Countries with above average number of companies certified with 27001 (2015)",
948   - x = reorder(country_short,X2015),
949   - y = X2015,
950   - xlab = "Country",
951   - ylab = "Number of certifications",
952   - data = Cert_PerCountry[Cert_PerCountry$X2015 > mean(Cert_PerCountry$X2015),]
953   - , geom = "col",
954   - fill = Continent)
955   - attacks2k15 <- Attacks[Attacks$Date < "2016-01-01" & Attacks$Date >= "2015-01-01",]
956   - frameAttacks2k15 <- as.data.frame(table(attacks2k15$Country))
957   - colnames(frameAttacks2k15) <- c("Country","Attacks")
958   - graph11 <- ggplot2::qplot(main = "Countries with above average number of cyberattacks (2015)",
959   - x = reorder(Country,Attacks),
960   - y = Attacks,
961   - xlab = "Country",
962   - ylab = "Number of attacks",
963   - data = frameAttacks2k15[frameAttacks2k15$Attacks > mean(frameAttacks2k15$Attacks),]
964   - , geom = "col",
965   - fill = Continent)
966   -
967   - Attacks2015ByMonth <- mutate(attacks2k15, month = format(attacks2k15$Date, "%m")) %>% group_by(month)
968   - Attack2015FreqByMonth <- as.data.frame(table(Attacks2015ByMonth$month))
969   - colnames(Attack2015FreqByMonth) <- c("Month", "Attacks")
970   - graph12 <- ggplot2::qplot(x = as.numeric(Month),
971   - y = Attacks,
972   - main = "Global cyberattack progression by month (2015)",
973   - data = Attack2015FreqByMonth,
974   - geom = c("point", "smooth"),
975   - xlim = c(1,12),
976   - xlab = "Month") + ggplot2::scale_x_continuous(breaks = 1:12)
977   -
978   -
979   -
980   -
981   - list(graph1,graph2,graph3,graph4,graph5,graph6,graph7,graph8,graph9,graph10,graph11,graph12)
982   -}
... ...
ISO27001effectiveness/Report.Rmd
1 1 ---
2   -title: "Estudio sobre la efectividad de la certificación ISO 27001"
  2 +title: "Estudio sobre la efectividad de la certificación ISO 27001"
3 3 output:
4 4 html_document: default
5 5 pdf_document: default
... ... @@ -199,193 +199,279 @@ Por otro lado tenemos _DNS_, cuya tendencia también tiene una pendiente negativ
199 199  
200 200 ### Evolución geográfica
201 201  
202   -Este apartado estudiará la relación entre la certificación ISO 27001 y los ataques producidos, pero teniendo en cuenta la variable geográfica, ya que es posible que la certificación, aunque sea internacional, se implemente de una mejor o peor forma según la región. En primer lugar se generalizará por continente.
  202 +Este apartado estudiará la relación entre la certificación ISO 27001 y los ataques producidos, pero teniendo en cuenta la variable geográfica, ya que es posible que la certificación, aunque sea internacional, se implemente o funcione de una mejor o peor forma según la región. En primer lugar se generalizará por continente.
203 203  
204 204 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
205 205 ContinentPies <- ISO27001effectiveness::GetContinentPie(Attacks,
206 206 c(-2.5, 0, 2, 0, -2.5),
207 207 c(1.4, 1, 0.5, 0.5, 0),
208 208 Cert_PerCountry,
209   -c(0, -1.5, -2.5, -2.5, -2.5),
210   -c(0.5, 0.2, 0.5, 1.8, -0.5))
  209 +c(-2.5, -2.5, 0, -1.5, -2.5),
  210 +c(-0.5, 1.9, 1, 0.25, 0.6))
211 211  
212   -ContinentPies[[1]]
  212 +ContinentPies[[2]]
213 213 ```
214 214 ```{r fig.width=4.5, fig.height=3}
215   -ContinentPies[[3]]
  215 +ContinentPies[[1]]
216 216 ```
217 217  
218   -Se puede observar a simple vista que los continentes que reciben más ataques, por una cuestión lógica de superficie e intereses, son por orden América, Asia y Europa. En cambio los continentes que mas certificaciones ISO 27001 obtienen son por orden Asia, Europa y América. Tanto África como Oceania podemos descartarlos en este estudio ya que sus porcentajes no son relevantes. Observemos ahora cómo influye esto en el tiempo.
  218 +Se puede observar a simple vista que los continentes que reciben más ataques, por lo que podría ser una cuestión lógica de superficie y/o intereses, son por orden: América, Asia y Europa. En cambio, los continentes que mas certificaciones ISO 27001 obtienen son los mismos, pero en distinto orden: Asia, Europa y América.
  219 +
  220 +Tanto África como Oceania serán descartados en el resto del estudio ya que sus porcentajes no parecen relevantes.
219 221  
220 222 ```{r fig.width=4.5, fig.height=4,out.extra='style="float:left"'}
221   -ISO27001effectiveness::GetContinentCertsEvolution(Cert_PerCountry)
  223 +ContinentCertEvolution <- ISO27001effectiveness::GetContinentCertsEvolution(Cert_PerCountry, ContinentPies[[3]])
  224 +
  225 +ContinentCertEvolution[[1]]
222 226 ```
223 227 ```{r fig.width=4.5, fig.height=4}
224   -ISO27001effectiveness::GetContinentAttacksEvolution(Attacks)
  228 +ContinentAttacksEvolution <- ISO27001effectiveness::GetContinentAttacksEvolution(Attacks, ContinentPies[[3]])
  229 +
  230 +ContinentAttacksEvolution[[1]]
225 231 ```
226 232  
227   -Se puede observar que la tendencia de las certificaciones es creciente mientras que la de los ataques es decreciente, pero procederemos a comparar cada una individualmente para poder demostrarlo numéricamente y no solo aparentemente.
  233 +Se puede observar que en conjunto la pendiente de la tendencia de las certificaciones es creciente, mientras que la de los ataques es decreciente. No obstante, se procederá a comparar cada continente individualmente para poder observarlo con más precisión.
228 234  
229 235 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
230   -topAttacks <- ISO27001effectiveness::GetContinentAttacksTopEvolution(Attacks)
231   -topCerts <- ISO27001effectiveness::GetContinentCertsTopEvolution(Cert_PerCountry)
232   -topCerts[[1]]
233   -slope_Ame_Cert <- (fit[5] - fit[1]) / 4
  236 +Ame_Cert <- GetContinentCertsSingleEvolution(ContinentCertEvolution[[2]], "Americas", "2013", 1400, 0.5)
  237 +Ame_Cert[[1]]
  238 +slope_Ame_Cert <- Ame_Cert[[2]]
234 239 ```
235 240 ```{r fig.width=4.5, fig.height=3}
236   -topAttacks[[1]]
237   -slope_Ame_Att <- (fit[5] - fit[1]) / 4
  241 +Ame_Att <- GetContinentAttacksSigleEvolution(ContinentAttacksEvolution[[2]], "Americas", "2014", 850, 0)
  242 +Ame_Att[[1]]
  243 +slope_Ame_Att <- Ame_Att[[2]]
238 244 ```
239 245  
240   -La pendiente de los ataques para _América_ es `r slope_Ame_Att` mientras que la pendiente de las certificaciones es `r slope_Ame_Cert`.
  246 +Para _América_, la pendiente de la tendencia en las certificaciones es `r slope_Ame_Cert`, mientras que la pendiente de los ataques es `r slope_Ame_Att`.
241 247  
242 248 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
243   -topCerts[[2]]
244   -slope_Asi_Cert <- (fit[5] - fit[1]) / 4
  249 +Asi_Cert <- GetContinentCertsSingleEvolution(ContinentCertEvolution[[2]], "Asia", "2013", 13500, 0.5)
  250 +Asi_Cert[[1]]
  251 +slope_Asi_Cert <- Asi_Cert[[2]]
245 252 ```
246 253 ```{r fig.width=4.5, fig.height=3}
247   -topAttacks[[2]]
248   -slope_Asi_Att <- (fit[5] - fit[1]) / 4
  254 +Asi_Att <- GetContinentAttacksSigleEvolution(ContinentAttacksEvolution[[2]], "Asia", "2014", 400, 0)
  255 +Asi_Att[[1]]
  256 +slope_Asi_Att <- Asi_Att[[2]]
249 257 ```
250 258  
251   -La pendiente de los ataques para _Asia_ es `r slope_Asi_Att` mientras que la pendiente de las certificaciones es `r slope_Asi_Cert`.
  259 +Para _Asia_, la pendiente de la tendencia en las certificaciones es `r slope_Asi_Cert`, mientras que la pendiente de los ataques es `r slope_Asi_Att`.
252 260  
253 261 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
254   -topCerts[[3]]
255   -slope_Eu_Cert <- (fit[5] - fit[1]) / 4
  262 +Eu_Cert <- GetContinentCertsSingleEvolution(ContinentCertEvolution[[2]], "Europe", "2013", 9500, 0.5)
  263 +Eu_Cert[[1]]
  264 +slope_Eu_Cert <- Eu_Cert[[2]]
256 265 ```
257 266 ```{r fig.width=4.5, fig.height=3}
258   -topAttacks[[3]]
259   -slope_Eu_Att <- (fit[5] - fit[1]) / 4
  267 +Eu_Att <- GetContinentAttacksSigleEvolution(ContinentAttacksEvolution[[2]], "Europe", "2014", 375, 0)
  268 +Eu_Att[[1]]
  269 +slope_Eu_Att <- Eu_Att[[2]]
260 270 ```
261 271  
262   -La pendiente de los ataques para _Europa_ es `r slope_Eu_Att` mientras que la pendiente de las certificaciones es `r slope_Eu_Cert`.
263   -
264   -Todas las pendientes podrían representar la relación que buscamos, en la que un aumento en las certificaciones produce un descenso en los ataques.
  272 +Para _Europa_, la pendiente de la tendencia en las certificaciones es `r slope_Eu_Cert`, mientras que la pendiente de los ataques es `r slope_Eu_Att`.
265 273  
266   -El análisis puede aumentar en profundidad estableciendo superficies geográficas más pequeñas y asi obtener más precisión, observemos lo que ocurre a nivel de paises. Se mostrarán a continuación los paises que superan aproximadamente un 2% de los ataques/certificaciones totales ya que la lista total de paises es demasiado extensa.
  274 +El análisis continuará estableciendo superficies geográficas más pequeñas, aumentando así la precisión, para poder observar lo que ocurre a nivel de paises. Se mostrarán a continuación los que superan aproximádamente un 2% de las certificaciones totales y el 1,5% para los ataques (porque _Estados Unidos_ recibe la gran mayoría y si no saldría solo), ya que la lista completa es demasiado extensa. Los nombres serán representados con los dos caracteres correspondientes al estándar ISO.
267 275  
268 276 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
269   -top <- GetCountriesCol(Attacks, Cert_PerCountry)
270   -top[[1]]
  277 +CountryCol <- GetCountriesCol(Attacks,0, 0.5,
  278 + Cert_PerCountry, 0, 0.5)
  279 +CountryCol[[2]]
271 280 ```
272 281 ```{r fig.width=4.5, fig.height=3}
273   -top[[2]]
  282 +CountryCol[[1]]
274 283 ```
275 284  
276   -Como se puede observar en la parte de certificaciones destaca de largo Japón sobre los demás, que se encuentra bastante bajo en la lista de ataques. Y al reves pasa algo parecido, en los ataques destaca Estados Unidos por mucho mientras que ese mismo pais está muy bajo en certificaciones. A continuación observaremos la evolución temporal del top 3 paises en ataques recibidos y en certificaciones obtenidas, varios de ellos coinciden, tenemos por la parte de las certificaciones a _Japón_, por la parte de los ataques a _Estados Unidos_, y común a ambas _Reino Unido_ e _India_.
  285 +Como se puede observar, en la parte de certificaciones destaca de largo _Japón_ sobre los demás, que a su vez se encuentra bastante bajo en la lista de ataques. Y, de un modo totalmente contrario, en los ataques destaca _Estados Unidos_ por mucho, mientras que este mismo está muy bajo en certificaciones.
  286 +
  287 +El estudio continuará en profundidad con el top 3 de paises en número de certificaciones y en número de ataques recibidos, aunque algunos de ellos coincidan.
277 288  
278 289 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
279   -topAttacks <- ISO27001effectiveness::GetCountriesAttacksTopEvolution(Attacks)
280   -topCerts <- ISO27001effectiveness::GetCountriesCertsTopEvolution(Cert_PerCountry)
281   -topCerts[[1]]
282   -slope_US_Cert <- (fit[5] - fit[1]) / 4
  290 +US_Cert <- GetCountriesCertsSingleEvolution(Cert_PerCountry,
  291 + CountryCol[[3]][1],
  292 + "2013", 875, 0.5)
  293 +
  294 +US_Cert[[1]]
  295 +slope_US_Cert <- US_Cert[[2]]
283 296 ```
284 297 ```{r fig.width=4.5, fig.height=3}
285   -topAttacks[[1]]
286   -slope_US_Att <- (fit[5] - fit[1]) / 4
  298 +US_Att <- GetCountriesAttacksSingleEvolution(Attacks,
  299 + CountryCol[[3]][1],
  300 + "2014", 725, 0)
  301 +
  302 +US_Att[[1]]
  303 +slope_US_Att <- US_Att[[2]]
287 304 ```
288 305  
289   -La pendiente de los ataques para _Estados Unidos_ es `r slope_US_Att` mientras que la pendiente de las certificaciones es `r slope_US_Cert`.
  306 +La pendiente de la tendencia en los ataques para _`r CountryCol[[3]][1]`_ es `r slope_US_Att` mientras que la pendiente de las certificaciones es `r slope_US_Cert`.
290 307  
291 308 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
292   -topCerts[[2]]
293   -slope_GB_Cert <- (fit[5] - fit[1]) / 4
  309 +GB_Cert <- GetCountriesCertsSingleEvolution(Cert_PerCountry,
  310 + CountryCol[[3]][2],
  311 + "2013", 2400, 0.5)
  312 +
  313 +GB_Cert[[1]]
  314 +slope_GB_Cert <- GB_Cert[[2]]
294 315 ```
295 316 ```{r fig.width=4.5, fig.height=3}
296   -topAttacks[[2]]
297   -slope_GB_Att <- (fit[5] - fit[1]) / 4
  317 +GB_Att <- GetCountriesAttacksSingleEvolution(Attacks,
  318 + CountryCol[[3]][2],
  319 + "2014", 125, 0.5)
  320 +
  321 +GB_Att[[1]]
  322 +slope_GB_Att <- GB_Att[[2]]
298 323 ```
299 324  
300   -La pendiente de los ataques para _Reino Unido_ es `r slope_GB_Att` mientras que la pendiente de las certificaciones es `r slope_GB_Cert`.
  325 +La pendiente de los ataques para _`r CountryCol[[3]][2]`_ es `r slope_GB_Att` mientras que la pendiente de las certificaciones es `r slope_GB_Cert`.
301 326  
302 327 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
303   -topCerts[[3]]
304   -slope_IN_Cert <- (fit[5] - fit[1]) / 4
  328 +IN_Cert <- GetCountriesCertsSingleEvolution(Cert_PerCountry,
  329 + CountryCol[[3]][3],
  330 + "2013", 1875, -0.2)
  331 +
  332 +IN_Cert[[1]]
  333 +slope_IN_Cert <- IN_Cert[[2]]
305 334 ```
306 335 ```{r fig.width=4.5, fig.height=3}
307   -topAttacks[[3]]
308   -slope_IN_Att <- (fit[5] - fit[1]) / 4
  336 +IN_Att <- GetCountriesAttacksSingleEvolution(Attacks,
  337 + CountryCol[[3]][3],
  338 + "2014", 60, -0.3)
  339 +
  340 +IN_Att[[1]]
  341 +slope_IN_Att <- IN_Att[[2]]
309 342 ```
310 343  
311   -La pendiente de los ataques para _India_ es `r slope_IN_Att` mientras que la pendiente de las certificaciones es `r slope_IN_Cert`.
  344 +La pendiente de los ataques para _`r CountryCol[[3]][3]`_ es `r slope_IN_Att` mientras que la pendiente de las certificaciones es `r slope_IN_Cert`.
312 345  
313 346 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
314   -topCerts[[4]]
315   -slope_JP_Cert <- (fit[5] - fit[1]) / 4
  347 +JP_Cert <- GetCountriesCertsSingleEvolution(Cert_PerCountry,
  348 + CountryCol[[3]][4],
  349 + "2013", 7600, 1)
  350 +
  351 +JP_Cert[[1]]
  352 +slope_JP_Cert <- JP_Cert[[2]]
316 353 ```
317 354 ```{r fig.width=4.5, fig.height=3}
318   -topAttacks[[4]]
319   -slope_JP_Att <- (fit[5] - fit[1]) / 4
  355 +JP_Att <- GetCountriesAttacksSingleEvolution(Attacks,
  356 + CountryCol[[3]][4],
  357 + "2014", 30, -0.5)
  358 +
  359 +JP_Att[[1]]
  360 +slope_JP_Att <- JP_Att[[2]]
320 361 ```
321 362  
322   -La pendiente de los ataques para _Japón_ es `r slope_JP_Att` mientras que la pendiente de las certificaciones es `r slope_JP_Cert`.
  363 +La pendiente de los ataques para _`r CountryCol[[3]][4]`_ es `r slope_JP_Att` mientras que la pendiente de las certificaciones es `r slope_JP_Cert`.
323 364  
324 365 ### Evolución geográfica y tipo de ataque
325 366  
326   -Hasta ahora se han analizado por separado estas dos variables, pero la respuesta podria encontrarse en una combinacion de las mismas. Para ello se analizarán los tipos de ataque reportados en el pais en el que parece que la certificación es más efectiva, _Japón_, y en el que menos, _Estados Unidos_.
  367 +Hasta ahora se han analizado por separado el tipo de ataque y la localización geofráfica, pero la respuesta podría encontrarse en una combinación de las mismas. Para ello se analizarán los tipos de ataque reportados en el país con más certificaciones, _Japón_, y en el que más ataques recibe, _Estados Unidos_.
  368 +
  369 +```{r fig.width=9, fig.height=4}
  370 +ISO27001effectiveness::GetContinentAttackCol(Attacks, "US", "JP")
  371 +```
  372 +
  373 +Como se pudo apreciar en un apartado previo, la ISO 27001 parece especialmente efectiva contra las técnicas de _Defacement_, _DDoS_ e _Injection_, en los gráficos previos se puede observar que _Estados Unidos_ tiene un menor porcentaje de este tipo de ataques con respecto a _Japón_. Para valorarlo mejor se representará a continuación cómo evolucionan con el tiempo en ambos paises.
327 374  
328 375 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
329   -graphs <- ISO27001effectiveness::GetContinentAttackPie(Attacks)
330   -graphs[[1]]
  376 +US_Att_Year <- ISO27001effectiveness::GetContinentAttackEvolution(Attacks, "US")
  377 +
  378 +US_Att_Year[[1]]
331 379 ```
332 380 ```{r fig.width=4.5, fig.height=3}
333   -graphs[[2]]
  381 +JP_Att_Year <- ISO27001effectiveness::GetContinentAttackEvolution(Attacks, "JP")
  382 +JP_Att_Year[[1]]
334 383 ```
335 384  
336   -Como vimos en el apartado previo, la ISO 27001 parece especialmente efectiva contra las técnicas de _Defacement_, _DDoS_ e _Injection_, en los gráficos previos podemos observar como para _Estados Unidos_ tiene un menor porcentaje de este tipo de ataques con respecto a _Japón_. Para valorarlo mejor se representará a continuación cómo evolucionan con el tiempo.
  385 +Como se puede observar, en _Estados Unidos_ se presenta una tendencia mayor a recibir los tipos de ataques que parecen menos afectados por la ISO 27001, mientras que en _Japón_ parecen matenerse. Se representarán las tendecias para cada técnica en ambos paises.
337 386  
338 387 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
339   -graphs <- ISO27001effectiveness::GetContinentAttackEvolution(Attacks)
340   -graphs[[1]]
  388 +US_dds_Year <- GetAttackTypeSigleEvolution(US_Att_Year[[2]],
  389 + "DDoS",
  390 + "2014",
  391 + 75, 0)
  392 +US_dds_Year[[1]]
  393 +slope_US_dds <- US_dds_Year[[2]]
341 394 ```
342 395 ```{r fig.width=4.5, fig.height=3}
343   -graphs[[2]]
  396 +JP_dds_Year <- GetAttackTypeSigleEvolution(JP_Att_Year[[2]],
  397 + "DDoS",
  398 + "2014",
  399 + 4.4, 0)
  400 +JP_dds_Year[[1]]
  401 +slope_JP_dds <- JP_dds_Year[[2]]
344 402 ```
345 403  
346   -Como se puede observar, en _Estados Unidos_ se presenta una tendencia mayor a recibir los tipos de ataques que parecen menos afectados por la ISO 27001, mientras que en _Japón_ parecen matenerse. Se representarán las tendecias en ambos paises.
  404 +Para _DDoS_, la pendiente de la tendencia es menor en _Estados Unidos_(`r slope_US_dds`) que en _Japón_(`r slope_JP_dds`).
347 405  
348 406 ```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
349   -graphs[[3]]
350   -slope_US_Oth <- (fit[5] - fit[1]) / 4
  407 +US_def_Year <- GetAttackTypeSigleEvolution(US_Att_Year[[2]],
  408 + "Defacement",
  409 + "2014",
  410 + 46, 0)
  411 +US_def_Year[[1]]
  412 +slope_US_def <- US_def_Year[[2]]
351 413 ```
352 414 ```{r fig.width=4.5, fig.height=3}
353   -graphs[[4]]
354   -slope_JP_Oth <- (fit[5] - fit[1]) / 4
  415 +JP_def_Year <- GetAttackTypeSigleEvolution(JP_Att_Year[[2]],
  416 + "Defacement",
  417 + "2014",
  418 + 1.5, 0)
  419 +JP_def_Year[[1]]
  420 +slope_JP_def <- JP_def_Year[[2]]
355 421 ```
356 422  
357   -Efectivamente, la tendencia en _Estados Unidos_ tiene una pendiente mayor (`r slope_US_Oth`) que en Japón que es cercana al 0 (`r slope_JP_Oth`) lo que, aunque la gráfica sea irregular debido a la baja cantidad de casos, implica una constancia.
358   -
359   -# Viejo
360   -
361   -De los datos mostrados se pueden hacer diferentes observaciones:
362   -
363   -* De 2014 a 2015, USA pasa de tener 654 a 1247 empresas con certificación ISO 27001, sin embargo la cifra de ciberataques se mantiene constante de 383 a 386 ataques recibidos. De 2013 a 2014 por ejemplo, pasa de recibir 505 a 383 ataques pese a sólo haber pasado de 566 a 654 empresas con dicha certificación.
364   -* Japón tiene un número inusualmente alto de empresas con la ISO 27001, no obstante sufre una cantidad de ciberataques comparativa a la de Israel, que tiene muchas menos empresas con la certificación.
365   -
366   -Ambas observaciones son de especial interés ya que ponen de relieve situaciones en las que de ser efectivo reformar los sistemas para cumplir la ISO 27001, debería poder apreciarse un efecto en la cantidad de ciberataques recibidos. En el caso de los USA, sólo podría explicarse mediante alguna de las siguientes hipótesis:
  423 +Para _Defacement_, la pendiente de la tendencia es menor en _Estados Unidos_(`r slope_US_def`) que en _Japón_(`r slope_JP_def`), aunque es muy parecida.
367 424  
368   -* Entre 2013 y 2014 se produjo un número especialmente alto de ciberataques a USA
369   -* Se crean más empresas de las que logran certificarse, y además, sufren ataques antes de obtenerla
  425 +```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
  426 +US_inj_Year <- GetAttackTypeSigleEvolution(US_Att_Year[[2]],
  427 + "Injection",
  428 + "2014",
  429 + 155, 0)
  430 +US_inj_Year[[1]]
  431 +slope_US_inj <- US_inj_Year[[2]]
  432 +```
  433 +```{r fig.width=4.5, fig.height=3}
  434 +JP_inj_Year <- GetAttackTypeSigleEvolution(JP_Att_Year[[2]],
  435 + "Injection",
  436 + "2014",
  437 + 9, 0)
  438 +JP_inj_Year[[1]]
  439 +slope_JP_inj <- JP_inj_Year[[2]]
  440 +```
370 441  
371   -La primera hipótesis puede ser comprobada con los datos que disponemos. A continuación se muestra una línea temporal de ciberataques globales entre los años 2012 y 2015.
  442 +Para _Injection_, la pendiente de la tendencia es menor en _Estados Unidos_(`r slope_US_inj`) que en _Japón_(`r slope_JP_inj`).
372 443  
  444 +```{r fig.width=4.5, fig.height=3,out.extra='style="float:left"'}
  445 +US_Oth_Year <- GetAttackTypeSigleEvolution(US_Att_Year[[2]],
  446 + "Otros",
  447 + "2014",
  448 + 100, 0)
  449 +US_Oth_Year[[1]]
  450 +slope_US_Oth <- US_Oth_Year[[2]]
  451 +```
  452 +```{r fig.width=4.5, fig.height=3}
  453 +JP_Oth_Year <- GetAttackTypeSigleEvolution(JP_Att_Year[[2]],
  454 + "Otros",
  455 + 3.5,
  456 + 3.5, 0)
  457 +JP_Oth_Year[[1]]
  458 +slope_JP_Oth <- JP_Oth_Year[[2]]
  459 +```
373 460  
374   -Observamos que se produjo lo contrario, bajó el número de ciberataques.
  461 +Para _Otros_, la pendiente de la tendencia es mayor en _Estados Unidos_(`r slope_US_inj`) que en _Japón_(`r slope_JP_inj`).
375 462  
376 463 ## Conclusiones
377   -Por lo observado anteriormente, se puede concluir que:
378 464  
379   -* La cantidad de ciberataques que recibe un país no se mitiga por el número de empresas que han obtenido la certificación ISO 270001
380 465  
381   -Esto no lleva a concluir que obtener la certificación ISO 27001 no es efectiva para reducir el número de ciberataques, sino que probablemente dependa de factores externos ajenos a este estudio
382 466  
383 467 ## Trabajo futuro
384   -Como trabajo futuro querríamos poder seguir en la línea de investigación acerca de los siguientes puntos:
  468 +Como trabajo futuro, el estudio podría seguir diversas líneas de investigación, resumidas en los siguientes puntos:
385 469  
386   -* ¿Qué factores producen un aumento o recesión en la cantidad de ciberataques recibidos?
387   -* ¿Qué sectores industriales reciben más ciberataques?
388   -* ¿Cuáles de esos sectores son los que más certificaciones obtienen?
  470 +* Mejorar la fuente de datos de ciberataques. Una fuente cuya administración, recopilación y mantenimiento no dependiera de una sola persona y fuera más constante, homogenea y detallada.
  471 +* Conectar los datos sobre web sites certificados por pais que nos provee la fuente de datos de certificaciones.
  472 +* Homogenizar los nombres y conectar los datos sobre sectores industriales que nos proveen ambas fuentes.
  473 +* Investigación más a fondo de los cambios realizados sobre la norma 27001, como la 27001:2013, para relacionarlos con las irregularidades en las curvas de ataques.
  474 +* Incorporar nuevas variables que puedan afectar al estudio, como parches importantes u otras normas parecidas.
389 475  
390 476 Creemos que la investigación de estas cuestiones puede dar más robustez a las conclusiones expuestas en este estudio.
391 477  
... ...
ISO27001effectiveness/Report.html
... ... @@ -11,7 +11,7 @@
11 11  
12 12  
13 13  
14   -<title>Estudio sobre la efectividad de la certificación ISO 27001</title>
  14 +<title>Estudio sobre la efectividad de la certificación ISO 27001</title>
15 15  
16 16 <script src="data:application/x-javascript;base64,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"></script>
17 17 <meta name="viewport" content="width=device-width, initial-scale=1" />
... ... @@ -115,7 +115,7 @@ $(document).ready(function () {
115 115  
116 116  
117 117  
118   -<h1 class="title toc-ignore">Estudio sobre la efectividad de la certificación ISO 27001</h1>
  118 +<h1 class="title toc-ignore">Estudio sobre la efectividad de la certificación ISO 27001</h1>
119 119  
120 120 </div>
121 121  
... ... @@ -372,74 +372,63 @@ $(document).ready(function () {
372 372 </div>
373 373 <div id="evolucion-geografica" class="section level3">
374 374 <h3>Evolución geográfica</h3>
375   -<p>Este apartado estudiará la relación entre la certificación ISO 27001 y los ataques producidos, pero teniendo en cuenta la variable geográfica, ya que es posible que la certificación, aunque sea internacional, se implemente de una mejor o peor forma según la región. En primer lugar se generalizará por continente.</p>
376   -<p><img src="data:image/png;base64,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" width="432" style="float:left" /> <img src="data:image/png;base64,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