In DIGI-VUB/text.alignment: Text Alignment with Smith-Waterman

knitr::opts_chunk$set(echo = TRUE, message = FALSE, comment = NA, eval = TRUE)

Smith Waterman

Smith-Waterman is an algorithm to identify similaries between sequences. The algorithm is explained in detail at https://en.wikipedia.org/wiki/Smith%E2%80%93Waterman_algorithm and finds a local optimal alignment between 2 sequences of letters.

This package implements the algorithm for sequences of letters as well as sequences of words and is usefull for text analytics researchers.

• The package uses similar code as the textreuse::local_align function and also allows to align character sequences next to aligning word sequences

Example usage

The package was set up in order to easily

• Find names in documents even if they are not correctly spelled
• Match 2 texts
• Find relevant sequences of texts in other texts

We show some examples of these use cases below.

library(text.alignment)

Example matching 2 names

a <- "Gaspard   Tournelly cardeur à laine"
b <- "Gaspard   Bourelly cordonnier"
smith_waterman(a, b)

a <- "Gaspard   T.  cardeur à laine"
b <- "Gaspard   Tournelly cardeur à laine"
smith_waterman(a, b, type = "characters")

Example matching 2 translations

a <- system.file(package = "text.alignment", "extdata", "example1.txt")
a <- readLines(a)
a <- paste(a, collapse = "\n")
b <- system.file(package = "text.alignment", "extdata", "example2.txt")
b <- readLines(b)
b <- paste(b, collapse = "\n")
cat(a, sep = "\n")
cat(b, sep = "\n")

smith_waterman(a, b, type = "words")

Find relevant sequences of texts in other texts

x <- smith_waterman("Lange rei", b)
x$b$tokens[x$b$alignment$from:x$b$alignment$to]
overview <- as.data.frame(x)
overview$b_from
overview$b_to
substr(overview$b, overview$b_from, overview$b_to)

Get alignment overview as a data.frame

x <- smith_waterman(a, b)
x <- as.data.frame(x, alignment_id = "matching-a-to-b")
str(x)

DIGI-VUB/text.alignment documentation built on Sept. 18, 2023, 7:26 a.m.