challenging-america-word-gap-prediction/run.py

from nltk.tokenize import word_tokenize
from nltk import trigrams
from collections import defaultdict, Counter
import pandas as pd
import csv


class GapPredictor:
    def __init__(self, alpha):
        self.model = defaultdict(lambda: defaultdict(lambda: 0))
        self.alpha = alpha
        self.vocab = set()
        self.DEFAULT_PREDICTION = "the:0.2 be:0.2 to:0.2 of:0.1 and:0.1 a:0.1 :0.1"

    @staticmethod
    def preprocess_text(text):
        text = text.lower().replace("-\\n", "").replace("\\n", " ")
        return text

    @staticmethod
    def _prepare_train_data():
        data = pd.read_csv(
            "train/in.tsv.xz",
            sep="\t",
            error_bad_lines=False,
            warn_bad_lines=False,
            header=None,
            quoting=csv.QUOTE_NONE,
            nrows=90000,
        )

        train_labels = pd.read_csv(
            "train/expected.tsv",
            sep="\t",
            error_bad_lines=False,
            header=None,
            quoting=csv.QUOTE_NONE,
            nrows=90000,
        )

        train_data = data[[6, 7]]
        train_data = pd.concat([train_data, train_labels], axis=1)
        train_data["final"] = train_data[6] + train_data[0] + train_data[7]

        return train_data

    def train_model(self):
        training_data = self._prepare_train_data()
        for index, row in training_data.iterrows():
            text = self.preprocess_text(str(row["final"]))
            words = word_tokenize(text)
            for w1, w2, w3 in trigrams(words, pad_right=True, pad_left=True):
                if w1 and w2 and w3:
                    self.model[(w2, w3)][w1] += 1
                    self.model[(w1, w2)][w3] += 1
                    self.vocab.add(w1)
                    self.vocab.add(w2)
                    self.vocab.add(w3)

        for word_pair in self.model:
            num_n_grams = float(sum(self.model[word_pair].values()))
            for word in self.model[word_pair]:
                self.model[word_pair][word] = (
                    self.model[word_pair][word] + self.alpha
                ) / (num_n_grams + self.alpha * len(self.vocab))

    def predict_probs(self, words):
        if len(words) < 3:
            return self.DEFAULT_PREDICTION

        word1, word2 = words[0], words[1]
        raw_prediction = dict(self.model[word1, word2])
        prediction = dict(Counter(raw_prediction).most_common(6))

        total_prob = 0.0
        str_prediction = ""

        for word, prob in prediction.items():
            total_prob += prob
            str_prediction += f"{word}:{prob} "

        if total_prob == 0.0:
            return self.DEFAULT_PREDICTION

        remaining_prob = 1 - total_prob

        if remaining_prob < 0.01:
            remaining_prob = 0.01

        str_prediction += f":{remaining_prob}"

        return str_prediction

    def prepare_output(self, input_file, output_file):
        with open(output_file, "w") as file:
            data = pd.read_csv(
                input_file,
                sep="\t",
                error_bad_lines=False,
                warn_bad_lines=False,
                header=None,
                quoting=csv.QUOTE_NONE,
            )
            for _, row in data.iterrows():
                text = self.preprocess_text(str(row[7]))
                words = word_tokenize(text)
                prediction = self.predict_probs(words)
                file.write(prediction + "\n")


predictor = GapPredictor(alpha=0.00002)
predictor.train_model()

predictor.prepare_output("dev-0/in.tsv.xz", "dev-0/out.tsv")
predictor.prepare_output("test-A/in.tsv.xz", "test-A/out.tsv")