An example on how to cleanup text for text classification in ML

preprocess_text.go

package main

import (
	"fmt"
	"regexp"
	"strings"
)

func preProcess(text string) string {
	// Find all chars that are not alphabet
	reg := regexp.MustCompile("[^a-zA-Z]+")

	// Replace those chars with spaces
	text = reg.ReplaceAllString(text, " ")

	text = strings.ToLower(text)

	// Tokenize on whitespace, while removing excess whitespace
	tokens := strings.Fields(text)

	// Join the tokens back to string
	return strings.Join(tokens, " ")
}

func main() {
	test := `
	The idea of designing a new language has been inspired by common dissatisfaction
	with C++. Go was announced in 2009 and in March 2012 version 1.0 was released.
	Go is used in Google’s production, as well as by many other companies
	and open-source projects.
	`

	r := preProcess(test)
	fmt.Println(r)
}

/*
OUTPUT:

the idea of designing a new language has been inspired by common dissatisfaction with c
go was announced in and in march version was released go is used in google s 
production as well as by many other companies and open source projects
*/