video_vector_generator.py 2.31 KB
import pandas as pd
import numpy as np
from gensim.models import Word2Vec
import activation as ac

BATCH_SIZE = 1000

def normalize_tag(tag):
    if isinstance(tag, str):
        new_tag = tag.lower().replace('[^a-zA-Z]', ' ')
        if new_tag.find(" (") != -1:
            new_tag = new_tag[:new_tag.find(" (")]
        new_tag = new_tag.replace(" ", "-")
        new_tag = new_tag.replace('"', "")
        return new_tag
    else:
        return tag
def vectorization_video():
    print('[0.1 0.2]')


if __name__ == '__main__':
    tag_vectors = Word2Vec.load("./tag_vectors.model").wv
    video_vectors = Word2Vec().wv   # Empty model

    # Load video recommendation tags.
    #video_tags = pd.read_csv('./segment_tags.csv', encoding='utf-8',error_bad_lines=False)
    video_tags = pd.read_csv('./videoIds.csv', encoding='utf-8',error_bad_lines=False)
    # Define batch variables.
    batch_video_ids = []
    batch_video_vectors = []
    error_tags = []

    for i, row in video_tags.iterrows():
        #print(row)
        video_id = row[0]
        video_vector = np.zeros(100)
        for segment_index in range(1, 6):
            #get tag and weight from here
            #use weight as input of non-linear function before this step
            #print("SEG TAG ",row[segment_index])
            tag, weight = row[segment_index].split(":")
            #print(tag)
            if tag in tag_vectors.vocab:
                video_vector = video_vector + (tag_vectors[tag] * float(weight))
            else:
                #print("unknown", tag)
                # Pass if tag is unknown
                if tag not in error_tags:
                    error_tags.append(tag)

        batch_video_ids.append(video_id)
        batch_video_vectors.append(video_vector)
        # Add video vectors.
        if (i+1) % BATCH_SIZE == 0:
            video_vectors.add(batch_video_ids, batch_video_vectors)
            batch_video_ids = []
            batch_video_vectors = []
            print("Video vectors created: ", i+1)

    # Add rest of video vectors.
    video_vectors.add(batch_video_ids, batch_video_vectors)
    print("error tags: ")
    print(error_tags)
    print(len(error_tags))

    video_vectors.save("videoId_vectors.model")

    # Usage
    # video_vectors = Word2Vec().wv.load("video_vectors.model")
    # video_vectors.most_similar("XwFj", topn=5)