s495731/KWT-2024
1
0
Fork 0
forked from bfijalkowski/KWT-2024
Code Pull Requests Activity

Compare commits

base: s495731:e343070e32a6f25a562eebc4fd75e5824724da00
Branches Tags
s495731:main
bfijalkowski:main
..
compare: s495731:71ca3b66ed4328dcefeefcfd415f5d7f1f41b52a
Branches Tags
s495731:main
bfijalkowski:main

No commits in common. "e343070e32a6f25a562eebc4fd75e5824724da00" and "71ca3b66ed4328dcefeefcfd415f5d7f1f41b52a" have entirely different histories.
e343070e32 ... 71ca3b66ed

2 changed files with 56 additions and 210 deletions
Show Stats Expand all files Collapse all files

182
lab/lab_01.ipynb
View File

@ -52,7 +52,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 2, "execution_count": 1,
"id": "narrow-romantic", "id": "narrow-romantic",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -71,7 +71,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 3, "execution_count": 2,
"id": "indonesian-electron", "id": "indonesian-electron",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -82,7 +82,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 4, "execution_count": 3,
"id": "compact-trinidad", "id": "compact-trinidad",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -92,7 +92,7 @@
"['Press the ENTER button']" "['Press the ENTER button']"
] ]
}, },
"execution_count": 4, "execution_count": 3,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -119,7 +119,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 5, "execution_count": 4,
"id": "exposed-daniel", "id": "exposed-daniel",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -139,7 +139,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 6, "execution_count": 5,
"id": "serial-velvet", "id": "serial-velvet",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -149,7 +149,7 @@
"['Press the ENTER button', 'Press the ENTER key']" "['Press the ENTER button', 'Press the ENTER key']"
] ]
}, },
"execution_count": 6, "execution_count": 5,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -176,7 +176,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 9, "execution_count": 6,
"id": "every-gibson", "id": "every-gibson",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -186,7 +186,7 @@
"[]" "[]"
] ]
}, },
"execution_count": 9, "execution_count": 6,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -213,26 +213,13 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 11, "execution_count": 7,
"id": "protected-rings", "id": "protected-rings",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"data": {
"text/plain": [
"['Press the ENTER button', 'Press the ENTER key']"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"def tm_lookup(sentence):\n", "def tm_lookup(sentence):\n",
" return [entry[1] for entry in translation_memory if entry[0].lower() == sentence.lower()]\n", " return ''"
"\n",
"tm_lookup('Wciśnij przycisk ENTER')"
] ]
}, },
{ {
@ -245,17 +232,17 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 8, "execution_count": 18,
"id": "severe-alloy", "id": "severe-alloy",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
{ {
"data": { "data": {
"text/plain": [ "text/plain": [
"[]" "''"
] ]
}, },
"execution_count": 8, "execution_count": 18,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -274,29 +261,13 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 16, "execution_count": 11,
"id": "structural-diesel", "id": "structural-diesel",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"data": {
"text/plain": [
"['Press the ENTER button', 'Press the ENTER key']"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"import string\n",
"\n",
"def tm_lookup(sentence):\n", "def tm_lookup(sentence):\n",
" sentence = sentence.translate(str.maketrans('', '', string.punctuation))\n", " return ''"
" return [entry[1] for entry in translation_memory if entry[0].lower() == sentence.lower()]\n",
"\n",
"tm_lookup('Wciśnij przycisk [ENTER]')"
] ]
}, },
{ {
@ -309,17 +280,17 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 17, "execution_count": 12,
"id": "brief-senegal", "id": "brief-senegal",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
{ {
"data": { "data": {
"text/plain": [ "text/plain": [
"[]" "''"
] ]
}, },
"execution_count": 17, "execution_count": 12,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -346,66 +317,13 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 26, "execution_count": 14,
"id": "mathematical-customs", "id": "mathematical-customs",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"name": "stdout",
"output_type": "stream",
"text": [
"Suggestion:\n"
]
},
{
"data": {
"text/plain": [
"['System restart required']"
]
},
"execution_count": 26,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"def differenceThanNotBiggerThanOneElement(firstSentence, secondSentence):\n",
" firstSentenceList = firstSentence.lower().split()\n",
" secondSentenceList = secondSentence.lower().split()\n",
"\n",
" diffNumber = 0\n",
"\n",
" for i in range(len(firstSentenceList)):\n",
" if(firstSentenceList[i] != secondSentenceList[i]):\n",
" diffNumber=diffNumber+1\n",
" if(diffNumber > 2):\n",
" return False\n",
"\n",
" return True\n",
"\n",
"def tm_lookup(sentence):\n", "def tm_lookup(sentence):\n",
" sentence = sentence.translate(str.maketrans('', '', string.punctuation))\n", " return ''"
"\n",
" exactMatchList = [entry[1] for entry in translation_memory if entry[0].lower() == sentence.lower()]\n",
"\n",
" if(len(exactMatchList) == 0):\n",
" diffMatchList = [entry[1] for entry in translation_memory if differenceThanNotBiggerThanOneElement(entry[0], sentence)]\n",
"\n",
" if(len(diffMatchList) > 0):\n",
" print('Suggestion:')\n",
" return diffMatchList\n",
"\n",
" else:\n",
" return exactMatchList\n",
"\n",
"\n",
"\n",
"\n",
"\n",
"\n",
" \n",
"\n",
"tm_lookup('Wymagane ponowne uruchomienie maszyny')"
] ]
}, },
{ {
@ -426,7 +344,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 27, "execution_count": 15,
"id": "humanitarian-wrong", "id": "humanitarian-wrong",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -444,7 +362,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 28, "execution_count": 16,
"id": "located-perception", "id": "located-perception",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -456,7 +374,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 29, "execution_count": 17,
"id": "advised-casting", "id": "advised-casting",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -466,7 +384,7 @@
"[('przycisk', 'button'), ('drukarka', 'printer')]" "[('przycisk', 'button'), ('drukarka', 'printer')]"
] ]
}, },
"execution_count": 29, "execution_count": 17,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -488,7 +406,7 @@
"id": "defensive-fifteen", "id": "defensive-fifteen",
"metadata": {}, "metadata": {},
"source": [ "source": [
"Odpowiedź: O(n * m)" "Odpowiedź:"
] ]
}, },
{ {
@ -501,27 +419,13 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 34, "execution_count": 19,
"id": "original-tunisia", "id": "original-tunisia",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"data": {
"text/plain": [
"[]"
]
},
"execution_count": 34,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"def glossary_lookup(sentence):\n", "def glossary_lookup(sentence):\n",
" sentence_words = sentence.lower().split()\n", " return ''"
" return [entry for entry in glossary if entry[0] in sentence_words]\n",
"\n",
"glossary_lookup('Każda Drukarka posiada Przycisk wznowienia drukowania')"
] ]
}, },
{ {
@ -534,27 +438,13 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 35, "execution_count": 20,
"id": "adolescent-semiconductor", "id": "adolescent-semiconductor",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"data": {
"text/plain": [
"[('przycisk', 'button'), ('drukarka', 'printer')]"
]
},
"execution_count": 35,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"def glossary_lookup(sentence):\n", "def glossary_lookup(sentence):\n",
" sentence_words = set(sentence.lower().split())\n", " return ''"
" return [entry for entry in glossary if entry[0] in sentence_words]\n",
"\n",
"glossary_lookup('Każda Drukarka posiada Przycisk wznowienia drukowania')"
] ]
} }
], ],
@ -577,7 +467,7 @@
"name": "python", "name": "python",
"nbconvert_exporter": "python", "nbconvert_exporter": "python",
"pygments_lexer": "ipython3", "pygments_lexer": "ipython3",
"version": "3.11.7" "version": "3.8.10"
}, },
"subtitle": "1. Podstawowe techniki wspomagania tłumaczenia", "subtitle": "1. Podstawowe techniki wspomagania tłumaczenia",
"title": "Komputerowe wspomaganie tłumaczenia", "title": "Komputerowe wspomaganie tłumaczenia",

84
lab/lab_02.ipynb
View File

@ -57,7 +57,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 22, "execution_count": 1,
"id": "confident-prison", "id": "confident-prison",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -80,44 +80,13 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 43, "execution_count": 2,
"id": "continental-submission", "id": "continental-submission",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"data": {
"text/plain": [
"['The printer is switched off',\n",
" 'Check the network settings',\n",
" 'System restart required']"
]
},
"execution_count": 43,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"\n",
"def ice_lookup(sentence, prev_sentence, next_sentence):\n", "def ice_lookup(sentence, prev_sentence, next_sentence):\n",
" for index in range(len(translation_memory)):\n", " return []"
" if index == 0:\n",
" continue\n",
" elif index + 1 >= len(translation_memory):\n",
" return []\n",
" else:\n",
" middleText = translation_memory[index]\n",
" prevText = translation_memory[index-1]\n",
" nextText = translation_memory[index+1]\n",
" if(sentence == middleText[0] and prev_sentence == prevText[0] and next_sentence == nextText[0]):\n",
" return [middleText[1], prevText[1], nextText[1]]\n",
" \n",
" return []\n",
" \n",
"\n",
" \n",
" \n",
"ice_lookup('Drukarka jest wyłączona','Sprawdź ustawienia sieciowe','Wymagane ponowne uruchomienie komputera') \n"
] ]
}, },
{ {
@ -150,7 +119,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 44, "execution_count": 3,
"id": "fourth-pillow", "id": "fourth-pillow",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -172,7 +141,7 @@
"id": "graduate-theorem", "id": "graduate-theorem",
"metadata": {}, "metadata": {},
"source": [ "source": [
"Odpowiedź: Tak, Spełnia warunki dla 1,2,3,4, " "Odpowiedź:"
] ]
}, },
{ {
@ -210,7 +179,7 @@
"id": "metallic-leave", "id": "metallic-leave",
"metadata": {}, "metadata": {},
"source": [ "source": [
"Odpowiedź: Nie, Spełnia dla warunku 3, poniewaz x = 4, y = 4, to wychodzi d(x,y) = 3, a nie 0" "Odpowiedź:"
] ]
}, },
{ {
@ -254,7 +223,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 45, "execution_count": 5,
"id": "secondary-wrist", "id": "secondary-wrist",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -264,7 +233,7 @@
"2" "2"
] ]
}, },
"execution_count": 45, "execution_count": 5,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -285,7 +254,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 46, "execution_count": 6,
"id": "associate-tuner", "id": "associate-tuner",
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
@ -304,7 +273,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 47, "execution_count": 7,
"id": "focal-pathology", "id": "focal-pathology",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -314,7 +283,7 @@
"0.9166666666666666" "0.9166666666666666"
] ]
}, },
"execution_count": 47, "execution_count": 7,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -325,7 +294,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 48, "execution_count": 8,
"id": "roman-ceiling", "id": "roman-ceiling",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -335,7 +304,7 @@
"0.9428571428571428" "0.9428571428571428"
] ]
}, },
"execution_count": 48, "execution_count": 8,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -346,7 +315,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 49, "execution_count": 9,
"id": "invisible-cambodia", "id": "invisible-cambodia",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
@ -356,7 +325,7 @@
"0.631578947368421" "0.631578947368421"
] ]
}, },
"execution_count": 49, "execution_count": 9,
"metadata": {}, "metadata": {},
"output_type": "execute_result" "output_type": "execute_result"
} }
@ -375,26 +344,13 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 51, "execution_count": 10,
"id": "genetic-cradle", "id": "genetic-cradle",
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"data": {
"text/plain": [
"['Press the ENTER button']"
]
},
"execution_count": 51,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [ "source": [
"def fuzzy_lookup(sentence, threshold):\n", "def fuzzy_lookup(sentence, threshold):\n",
" return [entry[1] for entry in translation_memory if levenshtein_similarity(entry[0],sentence ) > threshold]\n", " return []"
"\n",
"fuzzy_lookup('Wciśnij przycisk Enter', 0.5)"
] ]
} }
], ],
@ -417,7 +373,7 @@
"name": "python", "name": "python",
"nbconvert_exporter": "python", "nbconvert_exporter": "python",
"pygments_lexer": "ipython3", "pygments_lexer": "ipython3",
"version": "3.11.7" "version": "3.8.10"
}, },
"subtitle": "2. Zaawansowane użycie pamięci tłumaczeń", "subtitle": "2. Zaawansowane użycie pamięci tłumaczeń",
"title": "Komputerowe wspomaganie tłumaczenia", "title": "Komputerowe wspomaganie tłumaczenia",