Predictograph
by Deadpool
10th of December, 2018
What's a predictograph?
Predictograph is an algorithm that predicts the next word from the two words you provide. It's main features are:
- input can be written in any case, with or without punctuation
- numerics are not supposed to be provided as input or expected as output
- easy to use, fast, with elaborate explanation of functions
- allows weighted predictions for simpler predictive models
More on the underlying learning principle: https://en.wikipedia.org/wiki/Katz%27s_back-off_model
Inputs of the app
Line to be finished [text]: 3-word phrase that should be completed by the algorithm
Style [drop-down menu]: different users have different styles of communication, thus we formed different styles (formal, mixed, informal) to match our users' preferences
Simplicity penalization [slider]: predictions by simpler model tend to be less accurate, thus users can move the slider to make the correction more or less strict
The function that does most of the work
FindTrigram8 <- function(inp, data4, data3, data2, mutation = 75){
trig_a <- data.frame(ngram = character(), freq = numeric())
trig_b <- data.frame(ngram = character(), freq = numeric())
trig_c <- data.frame(ngram = character(), freq = numeric())
inp <- tolower(inp)
inp <- gsub("[[:punct:]]", "", inp)
inp2 <- gsub("_", " ", inp)
regex <- gsub(" ", "_", sprintf("%s%s%s", "^", inp2, "_"))
indices <- grep(regex, data4$feature)
if(length(indices) > 2) {
trig_a <- data4[indices, ]
trig_a$probability_percent <- round(((trig_a$frequency/sum(trig_a$frequency))*100), 2)
trig_a$target[1] <- strsplit(as.character(trig_a$feature), "_")[[1]][[4]]
trig_a$target[2] <- strsplit(as.character(trig_a$feature), "_")[[2]][[4]]
trig_a$target[3] <- strsplit(as.character(trig_a$feature), "_")[[3]][[4]]
trig_x2 <- trig_a[1:3, c("target", "probability_percent")]
} else {
y <- paste(strsplit(inp2," ")[[1]][c(2,3)], collapse=" ")
regex2 <- gsub(" ", "_", sprintf("%s%s%s", "^", y, "_"))
indices2 <- grep(regex2, data3$feature)
if(length(indices2) > 2){
trig_b <- data3[indices2, ]
trig_b$probability_percent <- round(((trig_b$frequency/sum(trig_b$frequency))*mutation), 2)
trig_b$target[1] <- strsplit(as.character(trig_b$feature), "_")[[1]][[3]]
trig_b$target[2] <- strsplit(as.character(trig_b$feature), "_")[[2]][[3]]
trig_b$target[3] <- strsplit(as.character(trig_b$feature), "_")[[3]][[3]]
trig_x2 <- trig_b[1:3, c("target", "probability_percent")]
} else {
y2 <- paste(strsplit(inp2," ")[[1]][3], collapse=" ")
regex3 <- gsub(" ", "_", sprintf("%s%s%s", "^", y2, "_"))
indices3 <- grep(regex3, data2$feature)
if(length(indices3) > 2) {
trig_c <- data2[indices3, ]
trig_c$probability_percent <- round(((trig_c$frequency/sum(trig_c$frequency))*(mutation*mutation/100)), 2)
trig_c$target[1] <- strsplit(as.character(trig_c$feature), "_")[[1]][[2]]
trig_c$target[2] <- strsplit(as.character(trig_c$feature), "_")[[2]][[2]]
trig_c$target[3] <- strsplit(as.character(trig_c$feature), "_")[[3]][[2]]
trig_x2 <- trig_c[1:3, c("target", "probability_percent")]
} else {
x <- c("the", "a", "to")
y <- c(1, 1, 1)
trig_x2 <- data.frame(target = x, probability_percent = y)
}
}
}
print(trig_x2)
}
Sample output
Find the app here: https://deadpool.shinyapps.io/Predictograph/