157 lines
4.1 KiB
Plaintext
157 lines
4.1 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 1,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"import nltk\n",
|
|
"import pandas as pd\n",
|
|
"import regex as re\n",
|
|
"from csv import QUOTE_NONE\n",
|
|
"from collections import Counter, defaultdict\n",
|
|
"\n",
|
|
"ENCODING = \"utf-8\""
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def clean_text(text):\n",
|
|
" res = str(text).lower().strip()\n",
|
|
" return res"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def get_csv(fname):\n",
|
|
" return pd.read_csv(\n",
|
|
" fname,\n",
|
|
" sep=\"\\t\",\n",
|
|
" on_bad_lines='skip',\n",
|
|
" header=None,\n",
|
|
" quoting=QUOTE_NONE,\n",
|
|
" encoding=ENCODING\n",
|
|
" )"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def train_model(data, model):\n",
|
|
" for _, row in data.iterrows():\n",
|
|
" words = nltk.word_tokenize(clean_text(row[607]))\n",
|
|
" for w1, w2 in nltk.bigrams(words, pad_left=True, pad_right=True):\n",
|
|
" if w1 and w2:\n",
|
|
" model[w2][w1] += 1\n",
|
|
" for w2 in model:\n",
|
|
" total_count = float(sum(model[w2].values()))\n",
|
|
" for w1 in model[w2]:\n",
|
|
" model[w2][w1] /= total_count"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def predict_data(read_path, save_path, model):\n",
|
|
" data = get_csv(read_path)\n",
|
|
"\n",
|
|
" with open(save_path, \"w\", encoding=ENCODING) as f:\n",
|
|
" for _, row in data.iterrows():\n",
|
|
" words = nltk.word_tokenize(clean_text(row[7]))\n",
|
|
" if len(words) < 3:\n",
|
|
" prediction = \"the:0.3 be:0.2 to:0.2 of:0.1 and:0.1 :0.1\"\n",
|
|
" else:\n",
|
|
" prediction = predict(words[0], model)\n",
|
|
" f.write(prediction + \"\\n\")\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def predict(word, model):\n",
|
|
" predictions = dict(model[word])\n",
|
|
" most_common = dict(Counter(predictions).most_common(6))\n",
|
|
"\n",
|
|
" total_prob = 0.0\n",
|
|
" str_prediction = \"\"\n",
|
|
"\n",
|
|
" for word, prob in most_common.items():\n",
|
|
" total_prob += prob\n",
|
|
" str_prediction += f\"{word}:{prob} \"\n",
|
|
"\n",
|
|
" if total_prob == 0.0:\n",
|
|
" return \"the:0.3 be:0.2 to:0.2 of:0.1 and:0.1 :0.1\"\n",
|
|
"\n",
|
|
" rem_prob = 1 - total_prob\n",
|
|
" if rem_prob < 0.01:\n",
|
|
" rem_prob = 0.01\n",
|
|
"\n",
|
|
" str_prediction += f\":{rem_prob}\"\n",
|
|
"\n",
|
|
" return str_prediction"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 2,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"data = get_csv(\"train/in.tsv.xz\")\n",
|
|
"\n",
|
|
"train_words = get_csv(\"train/expected.tsv\")\n",
|
|
"train_data = data[[6, 7]]\n",
|
|
"train_data = pd.concat([train_data, train_words], axis=1)\n",
|
|
"\n",
|
|
"train_data[607] = train_data[6] + train_data[0] + train_data[7]\n",
|
|
"\n",
|
|
"model = defaultdict(lambda: defaultdict(lambda: 0))\n",
|
|
"\n",
|
|
"train_model(train_data, model)\n",
|
|
"\n",
|
|
"predict_data(\"dev-0/in.tsv.xz\", \"dev-0/out.tsv\", model)\n",
|
|
"predict_data(\"test-A/in.tsv.xz\", \"test-A/out.tsv\", model)\n"
|
|
]
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "Python 3 (ipykernel)",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 3
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython3",
|
|
"version": "3.10.2"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 2
|
|
}
|