#csv #lemmatization #nlp #package #words #txt #lemmatizing

bin+lib rust_lemmatizer

A lemmatizing package for use with a .csv dictionary of lemmas and their corresponding words

3 releases (breaking)

0.3.0 Aug 16, 2022
0.2.0 Aug 16, 2022
0.1.0 Aug 12, 2022

#1572 in Text processing

GPL-3.0-or-later

12KB
182 lines

rust_lemmatizer

A lemmatizing package written in Rust.

How to Use

On Files

To lemmatize text in a file, run the get_lemmas script using the following arguments. This script saved to a file named "lemmatized.csv" or "lemmatized.txt" depending on user input.

Specify your filename, the path to your csv containing a list of lemmas and their associated forms, and your desired return format (csv, txt, or a Vec).

cargo run --bin get_lemmas [YOUR FILENAME] [PATH TO LEMMA FILE] [OUTPUT - txt, csv, or vec]

An example command would be:

cargo run --bin get_lemmas src/lemma_example.txt src/lemma_dict.csv csv 

On Strings

To lemmatize a string, run the get_lemmas_from_string script. Pass the following command, specifying your string, the file containing the list of lemmas and their forms, and your desired return format (csv, txt, or a Vec).

cargo run --bin get_lemmas_from_string ["YOUR STRING"] [PATH TO LEMMA FILE] [OUTPUT - txt, csv, or vec]

For example:

cargo run --bin get_lemmas_from_string "This is an example string." src/lemma_dict.csv vec

To use in your own project

For lemmatizing a file:

use rust_lemmatizer::get_words;

get_words(filename, dict_name, file_output);

For lemmatizing on a string:

use rust_lemmatizer::get_words_from_string;

get_words_from_string(string_to_analyze, dict_name, return_type);

Lemma List and Formatting

Source

The list of lemmas included in the GitHub repo was sourced from this repository by Lin Wei.

The list was created by referencing the British Nation Corpus (BNC), NodeBox Linguistics and Yasumasa Someya's lemma list. From the original repo:

This lemma list is provided "as is" and is free to use for any research and/or educational purposes. The list currently contains 186,523 words (tokens) in 84,487 lemma groups.

Formatting

To create your own list of lemmas for use with the library, create a csv file formatted like the one included here. Use two columns, the first containing your lemmas and the second containing comma-separated forms of the lemmas. Include a header for each column.

Dependencies

~1.2–1.6MB
~19K SLoC