dustalov

#!/bin/sh
CWD=$(basename "$PWD")
XZ_OPT="-T 0" tar --exclude '*~' -C ../ -cJvf "../$CWD.tar.xz" "$CWD"

dustalov

WATSET ?= ../watset-java/target/watset.jar
LCC ?= ../lcc

export LANG:=en_US.UTF-8
export LC_COLLATE:=C
export CLASSPATH := $(WATSET)

nodes:
	cut -f1,2 $(LCC)/eng_news_2016_10K/eng_news_2016_10K-co_s.txt | sed -re 's/\t/\n/g' | sort -u | wc -l
	cut -f1,2 $(LCC)/eng_news_2016_30K/eng_news_2016_30K-co_s.txt | sed -re 's/\t/\n/g' | sort -u | wc -l

dustalov

#!/usr/bin/env python3

__author__ = 'Dmitry Ustalov'
__credits__ = 'Sebastian Padó'
__license__ = 'MIT'

# This is an MIT-licensed implementation of the sigf toolkit for randomization tests:
# https://nlpado.de/~sebastian/software/sigf.shtml

import random

dustalov

#!/usr/bin/env groovy
import org.apache.commons.math3.stat.descriptive.moment.Mean
import org.apache.commons.math3.stat.descriptive.moment.StandardDeviation
import org.jgrapht.graph.SimpleWeightedGraph
import org.jgrapht.util.SupplierUtil
import org.nlpub.watset.graph.ChineseWhispers
import org.nlpub.watset.graph.NodeWeighting
import org.nlpub.watset.graph.MaxMax
import org.nlpub.watset.eval.Measurer
import org.nlpub.watset.graph.Watset

dustalov

LC_COLLATE = C
SEED = 1337
WCL_WRAPPER = /srv/definitions/wcl-extract

measure:
	./measure.py

kfold: wiki_really_all.txt
	./kfold.py --seed=$(SEED) $<

dustalov

#!/usr/bin/env python

# This script computes the normalized modified purity and inverse purity
# as according to this paper: https://aclweb.org/anthology/P14-1097.
# In fact, this program is currently quite a rough translation of
# the evaluation-verb-classes.perl script provided by Daisuke Kawahara.

import argparse
import re
import sys

dustalov

#!/usr/bin/awk -f
BEGIN {
    # significance level
    if (length(ALPHA) == 0) ALPHA = 0.05;

    # standard error estimation method: "basic" or "pooled"
    if (length(SE) == 0) SE = "basic";

    # one-tailed or two-tailed?
    if (TAILS != 2) TAILS = 1;

dustalov

#!/usr/bin/env groovy
import de.tudarmstadt.ukp.jwktl.JWKTL
import de.tudarmstadt.ukp.jwktl.api.filter.WiktionaryEntryFilter
import de.tudarmstadt.ukp.jwktl.api.util.Language

final languages = [en: Language.ENGLISH, ru: Language.RUSSIAN, de: Language.GERMAN]

if (args.length != 2 || !languages.containsKey(args[1] = args[1].toLowerCase())) {
    throw new IllegalArgumentException('Required arguments: <PARSED-WIKTIONARY> en|ru|de')
}

dustalov

#!/bin/bash -e
S=$(head -1)
CHARSETS=(utf8 cp1251 cp1252 koi8r koi8u iso-8859-5 maccyrillic)
for c1 in ${CHARSETS[*]}; do
for c2 in ${CHARSETS[*]}; do
for c3 in ${CHARSETS[*]}; do
for c4 in ${CHARSETS[*]}; do
echo -ne "$c1\t$c2\t$c3\t$c4\t"
<<<$S iconv -f=$c1 -t=$c2 -c | iconv -f=$c3 -t=$c4 -c
done

dustalov

#!/usr/bin/env ruby

require 'net/http'
require 'uri'
require 'nokogiri'

Example = Struct.new(:text, :source)

def ruscorpora(word)
  uri = URI('http://search.ruscorpora.ru/download-xml.xml')