Natural Language Processing

Using the data gathered from the Spotify API, we now want to extract and process the lyrics for each song. This is accomplished through scraping textual information, namely lyrical data, from the Genius Lyrics website. Following extraction, the lyrics are thoroughly processed and cleaned before undergoing sentiment analysis.

Scraping the Web

To get started, the script below imports lyricsgenius, a fundamental package libary allowing for web scraping of the Genius Lyrics website to retrieve the lyrics of any given song. Through the initialization of the genius variable, one can access the Genius API and retrieve the lyrics of any given song, such as “Too Many Nights” by Metro Boomin.

import lyricsgenius

genius = lyricsgenius.Genius("epFCxujgBe-Y6WrkZedI8kerKxiCpR6Rh0DAHYNlKDf9B4H1nXTdZIkj7krNUHVV")
song = genius.search_song("Too Many Nights", "Metro Boomin")

Searching for "Too Many Nights" by Metro Boomin...
Done.

First, we define a function that retrieves the lyrics for any song and artist from the Genius database. As shown below, it first searches for the track using the provided name and artist and then extracts the lyrics from the search results.

def get_song_lyrics(song_name, song_artist):
    song_genius = genius.search_song(song_name, song_artist)
    song_lyrics = song_genius.lyrics.partition("Lyrics")[2]
    # Remove any numbers followed by 'Embed'
    song_lyrics = re.sub(r"[\[].*?[\]]|\d+Embed", "", song_lyrics)
    # Remove text between square brackets
    song_lyrics = re.sub(r"(\-[A-Za-z]+\-)", "", song_lyrics)
    song_lyrics = re.sub(r'\d+', '', song_lyrics)

    return song_lyrics

Pre-Processing Text Data

Using the genius package, we define a function to fetch the lyrics of a song given the track name and artist. Once retrieved, the next step is to pre-process the textual data. This involves a cleansing process to eliminate profanity and patterns that may hinder the overall readability. The Python script contains the following steps:

1. Language Detection
2. Expanding Contractions
3. Converting Text to Lowercase
4. Spell Checking + Censoring
5. Removing Punctuations
6. Tokenization

The function detect_and_translate below is designed to identify and translate text into a specified language, specifically English. It first checks the language of the original text and compares it to the target language. If the detected language differs from the target language, the function utilizes GoogleTranslator to translate the input text into the target language (English).

# Function to detect and translate text
def detect_and_translate(track_lyrics, target_lang='en'):
    if detect(track_lyrics) == target_lang:
        return track_lyrics
    translator = GoogleTranslator(source='auto', target=target_lang)
    return translator.translate(track_lyrics)

We also develop various functions to support the preprocessing of textual data, streamlining the process and improving the accuracy of the final output. Among these functions are a method for removing punctuation from a given string of lyrics and a spell-checker that automatically finds and corrects any spelling errors.

def remove_punctuation(text):
    no_punct = ""
    for char in text:
        if char not in string.punctuation:
            no_punct = no_punct + char
    return no_punct  # return unpunctuated string

# Spell Check + Censor
spell = SpellChecker()

def spell_check(word_list_str):
    word_corrected_list = []
    for word in word_list_str.split():
        word_corrected = spell.correction(word)
        if word_corrected is not None:
            word_corrected_list.append(word_corrected)
        else:
            word_corrected_list.append(word)
    return word_corrected_list

The clean_song_lyrics function is designed to simplify the processing of lyrics for a specific song and artist. The function extracts the lyrics from the Genius database and performs a series of modifications, including expanding contractions, removing repetitive phrases, and converting the text to lowercase. It also ensures that the spelling is correct and eliminates any profanity. The end result is a cleaned set of lyrics, tokenized and encoded as a list of words.

def clean_song_lyrics(song_name, artist_name):
    genius_lyrics = get_song_lyrics(song_name, artist_name)
    lyrics_en = detect_and_translate(genius_lyrics, "en")
    
    no_contract = [contractions.fix(word) for word in lyrics_en.split()]
    no_contract_str = " ".join(no_contract).lower()  # lowercase
    no_contract_str = re.sub(r"nana|lala", "", no_contract_str)

    corrected = spell_check(no_contract_str) # <5> # Spell Check + Censor
    censored = profanity.censor(" ".join(corrected), censor_char="")
    no_punct = remove_punctuation(censored) # <6> # Remove Punctuation
    
    tokenized = word_tokenize(no_punct)  # Tokenize
    strencode = [i.encode("ascii", "ignore") for i in tokenized]  # Encode() method
    return [i.decode() for i in strencode]  # Decode() method

Removing Stop Words

We employ the Natural Language Toolkit (NLTK) library and its WordNetLemmatizer tool to filter out stopwords. By removing frequently used words like “the,” “and,” or “of,” the resulting text becomes more concise, enabling a more thorough examination of the lyrics and their underlying message.

def remove_stopwords_lyrics(clean_lyrics_decode):
    stopword = stopwords.words("english")
    stopword.extend(["yeah", "nanana", "nana", "oh", "la"])
    return [word for word in clean_lyrics_decode if word not in stopword]

Lemmatization

Next, we define a function to perform lemmatization on a set of words using the WordNetLemmatizer class from the NLTK library. Lemmatization helps to standardize words and reduce their complexity by reducing words to their root or base form. Our function specifically targets verbs and transforms different variations of the same verb into its most basic form.

from nltk.corpus import stopwords, wordnet

def get_wordnet_pos(tag):
    if tag.startswith("J"):
        return wordnet.ADJ
    elif tag.startswith("V"):
        return wordnet.VERB
    elif tag.startswith("N"):
        return wordnet.NOUN
    elif tag.startswith("R"):
        return wordnet.ADV
    else:
        return wordnet.NOUN

from nltk.tag import pos_tag
from nltk import pos_tag

def word_lemmatize(lyrics_cleaned):  # clean_lyrics_decode):
    pos_tags = pos_tag(lyrics_cleaned)
    wordnet_pos = [(word, get_wordnet_pos(pos_tag)) for (word, pos_tag) in pos_tags]

    wnl = WordNetLemmatizer()  # Lemmatize Lyrics
    return [wnl.lemmatize(word, tag) for word, tag in wordnet_pos]

In summary, the code above defines a function that makes use of the WordNetLemmatizer class from the NLTK library to conduct lemmatization specifically targeting verbs, thereby converting words to their most basic form.

---

Sentiment Analysis

Subsequently, the process involves the implementation of pipeline classes to carry out predictions using models accessible in the Hub. The code imports and employs multiple transformer models specifically designed for text classification and sentiment analysis. Specifically, the following procedure creates three distinct pipelines, each equipped with different models that facilitate the assessment of emotions and sentiment in textual content.

import transformers
from transformers import pipeline

# Initialize Genius API and sentiment classifiers
classifiers = [
    pipeline("text-classification", model='bhadresh-savani/distilbert-base-uncased-emotion', return_all_scores=True),
    pipeline("text-classification", model='cardiffnlp/twitter-roberta-base-sentiment', return_all_scores=True),
    pipeline("sentiment-analysis", return_all_scores=True)
]

No model was supplied, defaulted to distilbert-base-uncased-finetuned-sst-2-english and revision af0f99b (https://huggingface.co/distilbert-base-uncased-finetuned-sst-2-english).
Using a pipeline without specifying a model name and revision in production is not recommended.

The get_lyric_sentiment function takes in pre-processed lyrics as input and produces a dictionary of sentiment scores. It leverages three distinct classifiers to calculate the scores and aggregates them into a final result. For instance, one of these classifiers is the distilbert-base-uncased-emotion model, specifically trained to detect “emotions in texts such as sadness, joy, love, anger, fear, and surprise”.

# Function to perform sentiment analysis
def get_lyric_sentiment(lyrics, classifiers):
    text = ' '.join(lyrics)
    scores = {}
    for classifier in classifiers:
        try:
            predictions = classifier(text, truncation=True)
            for prediction in predictions[0]:
                scores[prediction['label']] = prediction['score']
        except Exception as e:
            print(f"Error during sentiment analysis: {e}")
    return scores

If the lyric sequence contains more than 512 tokens, it will trigger an error message indicating an exception encountered in the ‘embeddings’ layer. However, we have implemented measures to properly manage lyric sequences that exceed 512 words in the function mentioned above.

---

Putting it All Together

To summarize, the code efficiently collects data and performs text analysis on every song in a playlist. Specifically, it systematically processes a list of tracks and corresponding artists while simultaneously conducting a thorough cleaning procedure on the lyrics. The cleaning process involves removing all nonessential characters, resulting in a more precise depiction of the song’s content. The outcome is a comprehensive frequency analysis of each word in a song’s lyrics, providing deeper insights into the overall conveyed message.

Additionally, the program computes a sentiment score for each song based on the lyrics, indicating whether the lyrics are positive, negative, or neutral. It also collects information about the song and artist, such as the release date, length, popularity, and genre. Finally, the program compiles all this information into a dataframe for further analysis.

track_data = []
for i, track in all_tracks.iterrows():

    song_name = track["name"] #.partition(" (")[0]
    song_name = track['name'].partition(" (with")[0]
    song_name = song_name.partition(" - From")[0]
    
    artist_name = track["artist"]

    try:
        track_lyrics = clean_song_lyrics(song_name, artist_name)
        stopwords_removed = remove_stopwords_lyrics(track_lyrics)
        lemmatized = word_lemmatize(stopwords_removed)

        sentiment_scores = get_lyric_sentiment(stopwords_removed, classifiers)

        track_info = track.to_dict()
        track_info.update(sentiment_scores)

        track_info["lyrics"] = track_lyrics
        track_info["stopwords_removed"] = stopwords_removed
        track_info["lemmatized"] = lemmatized

        track_data.append(track_info)

    except Exception as e:
        print(f"Error processing track {track['name']} by {track['artist']}: {e}")

df_tracks = pd.DataFrame(track_data)

Searching for "Please Please Please" by Sabrina Carpenter...
Done.
Searching for "Si Antes Te Hubiera Conocido" by KAROL G...
Done.
Searching for "BIRDS OF A FEATHER" by Billie Eilish...
Done.
Searching for "Good Luck, Babe!" by Chappell Roan...
Done.
Searching for "A Bar Song (Tipsy)" by Shaboozey...
Done.
Searching for "Not Like Us" by Kendrick Lamar...
Done.
Error during sentiment analysis: The expanded size of the tensor (519) must match the existing size (514) at non-singleton dimension 1.  Target sizes: [1, 519].  Tensor sizes: [1, 514]
Searching for "MILLION DOLLAR BABY" by Tommy Richman...
Done.
Searching for "Too Sweet" by Hozier...
Done.
Searching for "Beautiful Things" by Benson Boone...
Done.
Searching for "I Had Some Help (Feat. Morgan Wallen)" by Post Malone...
Done.
Searching for "Espresso" by Sabrina Carpenter...
Done.
Searching for "i like the way you kiss me" by Artemas...
Done.
Searching for "Stargazing" by Myles Smith...
Done.
Searching for "LUNCH" by Billie Eilish...
Done.
Searching for "End of Beginning" by Djo...
Done.
Searching for "we can't be friends (wait for your love)" by Ariana Grande...
Done.
Searching for "Lose Control" by Teddy Swims...
Done.
Searching for "Tough" by Quavo...
Done.
Searching for "Austin" by Dasha...
Done.
Searching for "I Can Do It With a Broken Heart" by Taylor Swift...
Done.
Searching for "Houdini" by Eminem...
Done.
Searching for "Nasty" by Tinashe...
Done.
Searching for "Belong Together" by Mark Ambor...
Done.
Searching for "Slow It Down" by Benson Boone...
Done.
Searching for "HOT TO GO!" by Chappell Roan...
Done.
Searching for "GIRLS" by The Kid LAROI...
Done.
Searching for "greedy" by Tate McRae...
Done.
Searching for "Move" by Adam Port...
Done.
Searching for "Fortnight (feat. Post Malone)" by Taylor Swift...
Done.
Searching for "Saturn" by SZA...
Done.
Searching for "28" by Zach Bryan...
Done.
Searching for "Close To You" by Gracie Abrams...
Done.
Searching for "the boy is mine" by Ariana Grande...
Done.
Searching for "Stick Season" by Noah Kahan...
Done.
Searching for "I Don't Wanna Wait" by David Guetta...
Done.
Searching for "Smeraldo Garden Marching Band (feat. Loco)" by Jimin...
Done.
Searching for "Stumblin' In" by CYRIL...
Done.
Searching for "360" by Charli xcx...
Done.
Searching for "Rockstar" by LISA...
Done.
Searching for "One Of The Girls" by The Weeknd...
Done.
Searching for "Scared To Start" by Michael Marcagi...
Done.
Searching for "Lies Lies Lies" by Morgan Wallen...
Done.
Searching for "feelslikeimfallinginlove" by Coldplay...
Done.
Searching for "Parking Lot" by Mustard...
Done.
Searching for "Gata Only" by FloyyMenor...
Done.
Searching for "BAND4BAND (feat. Lil Baby)" by Central Cee...
Done.
Searching for "Santa" by Rvssian...
Done.
Searching for "Magnetic" by ILLIT...
Done.
Searching for "Water" by Tyla...
Done.
Searching for "Illusion" by Dua Lipa...
Done.

Asking to truncate to max_length but no maximum length is provided and the model has no predefined maximum length. Default to no truncation.

#df_tracks = pd.DataFrame(track_data)
df_tracks.to_csv("../assets/data/all_tracks+lyrics.csv", index=False)

df_tracks

	name	track_id	album	artist	artist_id	release_date	length	popularity	artist_pop	artist_genres	...	fear	surprise	LABEL_0	LABEL_1	LABEL_2	NEGATIVE	POSITIVE	lyrics	stopwords_removed	lemmatized
0	Please Please Please	5N3hjp1WNayUPZrA8kJmJP	Please Please Please	Sabrina Carpenter	74KM79TiuVKeVCqs8QtB0B	2024-06-06	186365	98	91	['pop']	...	0.000637	0.000841	0.251058	0.542962	0.205980	0.857851	0.142149	[i, know, i, have, good, judgment, i, know, i,...	[know, good, judgment, know, good, taste, funn...	[know, good, judgment, know, good, taste, funn...
1	Si Antes Te Hubiera Conocido	6WatFBLVB0x077xWeoVc2k	Si Antes Te Hubiera Conocido	KAROL G	790FomKkXshlbRYZFtlgla	2024-06-21	195824	91	89	['reggaeton', 'reggaeton colombiano', 'trap la...	...	0.001239	0.000147	0.030611	0.523545	0.445844	0.966250	0.033750	[what, what, we, are, ready, to, rule, summer,...	[ready, rule, summer, started, fire, would, me...	[ready, rule, summer, start, fire, would, meet...
2	BIRDS OF A FEATHER	6dOtVTDdiauQNBQEDOtlAB	HIT ME HARD AND SOFT	Billie Eilish	6qqNVTkY8uBg9cP3Jd7DAH	2024-05-17	210373	98	94	['art pop', 'pop']	...	0.434426	0.034063	0.122799	0.504202	0.372999	0.959893	0.040107	[i, want, you, to, stay, til, i, am, in, the, ...	[want, stay, til, grave, til, rot, away, dead,...	[want, stay, til, grave, til, rot, away, dead,...
3	Good Luck, Babe!	0WbMK4wrZ1wFSty9F7FCgu	Good Luck, Babe!	Chappell Roan	7GlBOeep6PqTfFi59PTUUN	2024-04-05	218423	94	86	['indie pop', 'pov: indie']	...	0.000427	0.000471	0.343443	0.554660	0.101897	0.981645	0.018355	[it, is, fine, it, is, cool, you, can, say, th...	[fine, cool, say, nothing, know, truth, guess,...	[fine, cool, say, nothing, know, truth, guess,...
4	A Bar Song (Tipsy)	2FQrifJ1N335Ljm3TjTVVf	A Bar Song (Tipsy)	Shaboozey	3y2cIKLjiOlp1Np37WiUdH	2024-04-12	171291	93	81	['pop rap']	...	0.009928	0.003751	0.057218	0.763994	0.178788	0.994249	0.005751	[my, baby, want, a, barking, she, is, been, te...	[baby, want, barking, telling, night, long, ga...	[baby, want, bark, tell, night, long, gasoline...
5	Not Like Us	6AI3ezQ4o3HUoP6Dhudph3	Not Like Us	Kendrick Lamar	2YZyLoL8N0Wb9xBt1NhZWg	2024-05-04	274192	96	92	['conscious hip hop', 'hip hop', 'rap', 'west ...	...	0.007257	0.004582	NaN	NaN	NaN	0.997331	0.002669	[psst, i, see, dead, people, mustard, on, the,...	[psst, see, dead, people, mustard, beat, musta...	[psst, see, dead, people, mustard, beat, musta...
6	MILLION DOLLAR BABY	7fzHQizxTqy8wTXwlrgPQQ	MILLION DOLLAR BABY	Tommy Richman	1WaFQSHVGZQJTbf0BdxdNo	2024-04-26	155151	86	83	['chill abstract hip hop']	...	0.032734	0.004189	0.280416	0.655696	0.063888	0.992852	0.007148	[do, it, baby, do, what, i, should, think, do,...	[baby, think, baby, could, think, baby, think,...	[baby, think, baby, could, think, baby, think,...
7	Too Sweet	4IadxL6BUymXlh8RCJJu7T	Unheard	Hozier	2FXC3k01G6Gw61bmprjgqS	2024-03-22	251424	83	85	['irish singer-songwriter', 'modern rock', 'po...	...	0.002293	0.001093	0.120370	0.648945	0.230685	0.984314	0.015686	[it, can, not, be, said, i, am, an, early, bir...	[said, early, bird, clock, say, word, baby, ne...	[say, early, bird, clock, say, word, baby, nev...
8	Beautiful Things	6tNQ70jh4OwmPGpYy6R2o9	Beautiful Things	Benson Boone	22wbnEMDvgVIAGdFeek6ET	2024-01-18	180304	91	85	['singer-songwriter pop']	...	0.996488	0.000843	0.033857	0.318134	0.648009	0.012351	0.987649	[for, a, while, there, it, was, rough, but, la...	[rough, lately, better, last, four, cold, reme...	[rough, lately, well, last, four, cold, rememb...
9	I Had Some Help (Feat. Morgan Wallen)	7221xIgOnuakPdLqT0F3nP	I Had Some Help	Post Malone	246dkjvS1zLTtiykXe5h60	2024-05-10	178205	95	90	['dfw rap', 'melodic rap', 'pop', 'rap']	...	0.000706	0.000557	0.210037	0.717910	0.072052	0.998456	0.001544	[you, got, a, got, ta, nerve, do, not, you, ba...	[got, got, ta, nerve, baby, hit, curb, made, t...	[get, get, ta, nerve, baby, hit, curb, make, t...
10	Espresso	2qSkIjg1o9h3YT9RAgYN75	Espresso	Sabrina Carpenter	74KM79TiuVKeVCqs8QtB0B	2024-04-12	175459	99	91	['pop']	...	0.027107	0.004054	0.070886	0.800304	0.128810	0.992557	0.007443	[now, he, is, thinking, about, me, every, nigh...	[thinking, every, night, sweet, guess, say, sl...	[think, every, night, sweet, guess, say, sleep...
11	i like the way you kiss me	2GxrNKugF82CnoRFbQfzPf	i like the way you kiss me	Artemas	0PCCGZ0wGLizHt2KZ7hhA2	2024-03-19	142514	95	81	[]	...	0.001377	0.001145	0.034661	0.521730	0.443609	0.984495	0.015505	[i, like, the, way, you, kiss, me, i, like, th...	[like, way, kiss, like, way, uh, like, way, ki...	[like, way, kiss, like, way, uh, like, way, ki...
12	Stargazing	3Vr3zh0r7ALn8VLqCiRR10	Stargazing	Myles Smith	3bO19AOone0ubCsfDXDtYt	2024-05-10	172533	92	78	['singer-songwriter pop']	...	0.000203	0.000153	0.479969	0.450779	0.069251	0.994339	0.005661	[oohooh, oohooh, oohooh, time, stood, still, j...	[oohooh, oohooh, oohooh, time, stood, still, l...	[oohooh, oohooh, oohooh, time, stand, still, l...
13	LUNCH	629DixmZGHc7ILtEntuiWE	HIT ME HARD AND SOFT	Billie Eilish	6qqNVTkY8uBg9cP3Jd7DAH	2024-05-17	179586	94	94	['art pop', 'pop']	...	0.004807	0.005700	0.226971	0.703515	0.069515	0.995480	0.004520	[oh, mmmm, i, could, eat, that, girl, for, lun...	[mmmm, could, eat, girl, lunch, dances, tongue...	[mmmm, could, eat, girl, lunch, dance, tongue,...
14	End of Beginning	3qhlB30KknSejmIvZZLjOD	DECIDE	Djo	5p9HO3XC5P3BLxJs5Mtrhm	2022-09-16	159245	93	77	['pov: indie', 'psychedelic pop']	...	0.018631	0.004027	0.159378	0.691760	0.148862	0.964887	0.035113	[just, one, more, tear, to, cry, one, teardrop...	[one, tear, cry, one, teardrop, eye, better, s...	[one, tear, cry, one, teardrop, eye, well, sav...
15	we can't be friends (wait for your love)	46kspZSY3aKmwQe7O77fCC	eternal sunshine	Ariana Grande	66CXWjxzNUsdJxJ2JdwvnR	2024-03-08	228639	86	91	['pop']	...	0.000605	0.000460	0.063339	0.507926	0.428735	0.996296	0.003704	[i, did, not, think, you, would, understand, m...	[think, would, understand, could, ever, even, ...	[think, would, understand, could, ever, even, ...
16	Lose Control	17phhZDn6oGtzMe56NuWvj	I've Tried Everything But Therapy (Part 1)	Teddy Swims	33qOK5uJ8AR2xuQQAhHump	2023-09-15	210688	90	82	[]	...	0.000485	0.000120	0.557471	0.412347	0.030183	0.998184	0.001816	[something, is, got, a, hold, of, me, lately, ...	[something, got, hold, lately, know, anymore, ...	[something, get, hold, lately, know, anymore, ...
17	Tough	22DH8NChecsgPxDjA4pqer	Tough	Quavo	0VRj0yCOv2FXJNP47XQnx5	2024-07-03	188828	85	79	['atl hip hop', 'melodic rap', 'rap', 'trap']	...	0.041581	0.004940	0.152115	0.734467	0.113419	0.977612	0.022388	[tough, like, the, scuff, on, a, pair, of, old...	[tough, like, scuff, pair, old, leather, boots...	[tough, like, scuff, pair, old, leather, boot,...
18	Austin	4NJqhmkGN042BrvHoMKUrJ	Austin	Dasha	7Ez6lTtSMjMf2YSYpukP1I	2023-11-17	171782	41	75	[]	...	0.002042	0.000548	0.332335	0.619888	0.047777	0.997655	0.002345	[we, had, a, plan, move, out, of, this, town, ...	[plan, move, town, baby, west, sand, talked, l...	[plan, move, town, baby, west, sand, talk, lat...
19	I Can Do It With a Broken Heart	4q5YezDOIPcoLr8R81x9qy	THE TORTURED POETS DEPARTMENT	Taylor Swift	06HL4z0CvFAxyc27GXpf02	2024-04-18	218004	87	100	['pop']	...	0.001501	0.000734	0.268903	0.570628	0.160469	0.992031	0.007969	[i, can, read, your, mind, she, is, having, th...	[read, mind, time, life, glittering, prime, li...	[read, mind, time, life, glitter, prime, light...
20	Houdini	2HYFX63wP3otVIvopRS99Z	Houdini	Eminem	7dGJo4pcD2V6oG8kP0tJRR	2024-05-31	227239	94	91	['detroit hip hop', 'hip hop', 'rap']	...	0.059232	0.007880	0.143081	0.793169	0.063750	0.994274	0.005726	[hey, them, it, is, pal, uh, i, was, listening...	[hey, pal, uh, listening, album, good, lucky, ...	[hey, pal, uh, listen, album, good, lucky, gue...
21	Nasty	6NjWCIYu1W8xa3HIvcIhd4	Nasty	Tinashe	0NIIxcxNHmOoyBx03SfTCD	2024-04-12	176027	86	73	['alternative r&b', 'dance pop', 'metropopolis...	...	0.003332	0.002755	0.546907	0.416473	0.036621	0.997654	0.002346	[because, it, feels, like, heaven, when, it, h...	[feels, like, heaven, hurts, bad, baby, put, l...	[feel, like, heaven, hurt, bad, baby, put, lik...
22	Belong Together	5uQ7de4EWjb3rkcFxyEOpu	Belong Together	Mark Ambor	11p2E654TTU8e0nZWBR4AL	2024-02-16	148317	91	76	['singer-songwriter pop']	...	0.005112	0.002070	0.018785	0.284418	0.696797	0.206854	0.793146	[i, know, sleep, is, friends, with, death, but...	[know, sleep, friends, death, maybe, get, rest...	[know, sleep, friend, death, maybe, get, rest,...
23	Slow It Down	51eSHglvG1RJXtL3qI5trr	Fireworks & Rollerblades	Benson Boone	22wbnEMDvgVIAGdFeek6ET	2024-04-05	161831	89	85	['singer-songwriter pop']	...	0.669380	0.007431	0.709808	0.263808	0.026384	0.983390	0.016610	[i, would, never, met, you, but, i, wanted, to...	[would, never, met, wanted, invite, party, wal...	[would, never, meet, want, invite, party, walk...
24	HOT TO GO!	4xdBrk0nFZaP54vvZj0yx7	The Rise and Fall of a Midwest Princess	Chappell Roan	7GlBOeep6PqTfFi59PTUUN	2023-09-22	184841	88	86	['indie pop', 'pov: indie']	...	0.001401	0.002167	0.105176	0.805817	0.089007	0.988687	0.011313	[five, six, five, six, seven, eight, i, could,...	[five, six, five, six, seven, eight, could, on...	[five, six, five, six, seven, eight, could, on...
25	GIRLS	7z3PblAN3dH1JMewiRydkZ	GIRLS	The Kid LAROI	2tIP7SsRs7vjIcLrU85W8J	2024-06-28	152979	82	81	['australian hip hop']	...	0.004544	0.001163	0.171786	0.707467	0.120747	0.994324	0.005676	[fall, in, love, for, no, reason, fallin, uh, ...	[fall, love, reason, fallin, uh, turned, said,...	[fall, love, reason, fallin, uh, turn, say, gi...
26	greedy	3rUGC1vUpkDG9CZFHMur1t	greedy	Tate McRae	45dkTj5sMRSjrmBSBeiHym	2023-09-15	131872	89	82	['pop']	...	0.001383	0.000401	0.257882	0.650690	0.091428	0.970451	0.029549	[wood, he, said, are, you, serious, i, have, t...	[wood, said, serious, tried, figure, next, nig...	[wood, say, serious, tried, figure, next, nigh...
27	Move	1BJJbSX6muJVF2AK7uH1x4	Move	Adam Port	2loEsOijJ6XiGzWYFXMIRk	2024-06-07	177598	86	72	['melodic house']	...	0.001471	0.001813	0.125568	0.807287	0.067145	0.155766	0.844234	[fire, burning, style, gunning, any, i, am, fe...	[fire, burning, style, gunning, feenin, want, ...	[fire, burning, style, gun, feenin, want, tast...
28	Fortnight (feat. Post Malone)	2OzhQlSqBEmt7hmkYxfT6m	THE TORTURED POETS DEPARTMENT	Taylor Swift	06HL4z0CvFAxyc27GXpf02	2024-04-18	228965	91	100	['pop']	...	0.097827	0.012489	0.494738	0.437140	0.068122	0.984428	0.015572	[i, should, have, been, taken, away, but, peop...	[taken, away, people, forgot, come, pick, alco...	[take, away, people, forget, come, pick, alcoh...
29	Saturn	1bjeWoagtHmUKputLVyDxQ	Saturn	SZA	7tYKF4w9nC0nq9CsPZTHyP	2024-02-22	186191	90	89	['pop', 'r&b', 'rap']	...	0.077055	0.002667	0.401661	0.494084	0.104256	0.998026	0.001974	[if, there, is, another, universe, please, mak...	[another, universe, please, make, noise, noise...	[another, universe, please, make, noise, noise...
30	28	5iJKGpnFfvbjZJeAtwXfCj	The Great American Bar Scene	Zach Bryan	40ZNYROS4zLfyyBSs2PGe2	2024-07-04	233333	81	91	['classic oklahoma country']	...	0.000493	0.000980	0.127071	0.535515	0.337414	0.918328	0.081672	[you, took, a, train, to, the, south, side, of...	[took, train, south, side, boson, showed, old,...	[take, train, south, side, boson, show, old, m...
31	Close To You	5MPi9e7z46wopyad10R6qx	Close To You	Gracie Abrams	4tuJ0bMpJh08umKkEXKUI5	2024-06-07	225973	77	82	['alt z']	...	0.080101	0.008835	0.207858	0.741581	0.050560	0.963049	0.036952	[close, to, you, close, to, you, i, do, not, g...	[close, close, got, single, problem, provocati...	[close, close, get, single, problem, provocati...
32	the boy is mine	0Lmbke3KNVFXtoH2mMSHCw	eternal sunshine	Ariana Grande	66CXWjxzNUsdJxJ2JdwvnR	2024-03-08	173639	86	91	['pop']	...	0.009534	0.002129	0.181925	0.762790	0.055285	0.988922	0.011078	[how, can, it, be, you, and, me, might, be, me...	[might, meant, unseen, want, scene, usually, u...	[might, mean, unseen, want, scene, usually, un...
33	Stick Season	0mflMxspEfB0VbI1kyLiAv	Stick Season	Noah Kahan	2RQXRUsr4IW1f3mKyKsy4B	2022-10-14	182346	90	84	['pov: indie']	...	0.104740	0.003200	0.516024	0.447031	0.036945	0.998780	0.001220	[as, you, promised, me, that, i, was, more, th...	[promised, miles, combined, must, change, hear...	[promise, mile, combine, must, change, heart, ...
34	I Don't Wanna Wait	331l3xABO0HMr1Kkyh2LZq	I Don't Wanna Wait	David Guetta	1Cs0zKBU1kc0i8ypK3B9ai	2024-04-05	149667	90	89	['big room', 'dance pop', 'edm', 'pop', 'pop d...	...	0.000750	0.001233	0.186699	0.667073	0.146227	0.926766	0.073234	[let, us, make, tonight, the, weekend, i, do, ...	[let, us, make, tonight, weekend, want, waiait...	[let, u, make, tonight, weekend, want, waiait,...
35	Smeraldo Garden Marching Band (feat. Loco)	1kPhV0KQui1phEpjnWIqUN	Smeraldo Garden Marching Band (feat. Loco)	Jimin	1oSPZhvZMIrWW5I41kPkkY	2024-06-28	182840	91	84	['k-pop']	...	0.002519	0.001851	0.053019	0.700006	0.246975	0.068059	0.931941	[say, oh, this, harmonys, just, for, you, oh, ...	[say, harmonys, say, love, introducing, began,...	[say, harmony, say, love, introduce, begin, tu...
36	Stumblin' In	0h3Xy4V4apMraB5NuM8U7Z	Stumblin' In	CYRIL	11kt6ggsdxvI8MhyeSMKom	2023-11-10	213363	89	79	[]	...	0.327841	0.003080	0.325368	0.583619	0.091013	0.995401	0.004599	[our, love, is, alive, and, so, we, begin, foo...	[love, alive, begin, foolishly, laying, hearts...	[love, alive, begin, foolishly, lay, heart, ta...
37	360	4w2GLmK2wnioVnb5CPQeex	BRAT	Charli xcx	25uiPmTg16RbhZWAqwLBy5	2024-06-07	133805	84	85	['art pop', 'candy pop', 'metropopolis', 'pop'...	...	0.012352	0.004491	0.165593	0.730096	0.104311	0.991431	0.008569	[i, went, my, own, way, and, i, made, it, i, a...	[went, way, made, favorite, reference, baby, c...	[go, way, make, favorite, reference, baby, cal...
38	Rockstar	6vvPecFTmWxDfEJ6cYT1wa	Rockstar	LISA	5L1lO4eRHmJ7a0Q6csE5cT	2024-06-27	138213	90	79	['k-pop']	...	0.036838	0.012582	0.055340	0.774254	0.170406	0.981867	0.018133	[gold, teeth, sitting, on, the, dash, she, a, ...	[gold, teeth, sitting, dash, rockstar, make, f...	[gold, teeth, sit, dash, rockstar, make, favor...
39	One Of The Girls (with JENNIE, Lily Rose Depp)	7CyPwkp0oE8Ro9Dd5CUDjW	The Idol Episode 4 (Music from the HBO Origina...	The Weeknd	1Xyo4u8uXC1ZmMpatF05PJ	2023-06-23	244684	91	93	['canadian contemporary r&b', 'canadian pop', ...	...	0.018034	0.003468	0.178038	0.707149	0.114813	0.885132	0.114868	[lock, me, up, and, throw, away, the, key, he,...	[lock, throw, away, key, knows, get, best, for...	[lock, throw, away, key, know, get, best, forc...
40	Scared To Start	3Pbp7cUCx4d3OAkZSCoNvn	Scared To Start	Michael Marcagi	4j96cMcT8GRi11qbvo1cLQ	2024-01-12	159636	88	73	[]	...	0.993651	0.000680	0.467322	0.498693	0.033985	0.993110	0.006890	[she, is, wearing, an, old, dress, walking, it...	[wearing, old, dress, walking, waiting, someon...	[wear, old, dress, walk, wait, someone, turn, ...
41	Lies Lies Lies	7Fzl7QaTu47WyP9R5S5mh5	Lies Lies Lies	Morgan Wallen	4oUHIQIBe0LHzYfvXNW4QM	2024-07-05	198068	80	90	['contemporary country']	...	0.005516	0.000481	0.621788	0.354477	0.023735	0.990984	0.009016	[i, do, not, come, down, with, the, sun, i, wi...	[come, sun, hate, morning, comes, thoughts, bo...	[come, sun, hate, morning, come, thought, body...
42	feelslikeimfallinginlove	1YsU8rW2u8z4F0pwOBQ4Ea	feelslikeimfallinginlove	Coldplay	4gzpq5DPGxSnKTe4SA8HAU	2024-06-21	237803	83	88	['permanent wave', 'pop']	...	0.001050	0.000178	0.352217	0.564125	0.083659	0.954148	0.045852	[i, know, that, this, could, hurt, me, bad, i,...	[know, could, hurt, bad, know, could, feel, li...	[know, could, hurt, bad, know, could, feel, li...
43	Parking Lot	4IFd7EVCyJsUHesBMXI8ju	Parking Lot	Mustard	0YinUQ50QDB7ZxSCLyQ40k	2024-06-21	172794	83	73	['cali rap', 'pop rap', 'rap', 'southern hip h...	...	0.091244	0.010612	0.097719	0.778541	0.123740	0.991387	0.008613	[away, away, away, away, away, away, away, awa...	[away, away, away, away, away, away, away, awa...	[away, away, away, away, away, away, away, awa...
44	Gata Only	6XjDF6nds4DE2BBbagZol6	Gata Only	FloyyMenor	7CvTknweLr9feJtRGrpDBy	2024-02-02	222000	96	82	['reggaeton chileno']	...	0.884537	0.003583	0.210103	0.690181	0.099717	0.996415	0.003585	[hey, mommy, i, feel, you, far, away, tell, me...	[hey, mommy, feel, far, away, tell, want, goin...	[hey, mommy, feel, far, away, tell, want, go, ...
45	BAND4BAND (feat. Lil Baby)	7iabz12vAuVQYyekFIWJxD	BAND4BAND (feat. Lil Baby)	Central Cee	5H4yInM5zmHqpKIoMNAx4r	2024-05-23	140733	91	82	['melodic drill', 'r&drill']	...	0.031337	0.003643	0.158056	0.726879	0.115065	0.995000	0.005000	[i, am, not, in, the, mood, because, my, fligh...	[mood, flight, delayed, jumped, private, jet, ...	[mood, flight, delay, jump, private, jet, ask,...
46	Santa	5bi0gh89wRuH2OgjdAKFsb	Santa	Rvssian	1fctva4kpRbg2k3v7kwRuS	2024-04-04	193038	88	76	['reggaeton', 'reggaeton flow', 'trap latino',...	...	0.221692	0.006951	0.261035	0.679751	0.059214	0.986848	0.013152	[they, rvssian, i, can, not, deny, this, desir...	[rvssian, deny, desire, stop, looking, lips, r...	[rvssian, deny, desire, stop, look, lip, rejec...
47	Magnetic	1aKvZDoLGkNMxoRYgkckZG	SUPER REAL ME	ILLIT	36cgvBn0aadzOijnjjwqMN	2024-03-25	160688	89	74	['5th gen k-pop']	...	0.069957	0.007273	0.092965	0.713226	0.193809	0.986112	0.013888	[baby, i, am, just, trying, to, play, it, cool...	[baby, trying, play, cool, hide, want, wait, m...	[baby, try, play, cool, hide, want, wait, minu...
48	Water	5aIVCx5tnk0ntmdiinnYvw	Water	Tyla	3SozjO3Lat463tQICI9LcE	2023-07-28	200255	86	77	[]	...	0.006878	0.002168	0.210546	0.642083	0.147372	0.992614	0.007386	[make, me, sweat, make, me, hotter, make, me, ...	[make, sweat, make, hotter, make, lose, breath...	[make, sweat, make, hotter, make, lose, breath...
49	Illusion	59xD5osEFsaNt5PXfIKUnX	Illusion	Dua Lipa	6M2wZ9GZgrQXHCFfjv46we	2024-04-11	188143	79	88	['dance pop', 'pop', 'uk pop']	...	0.292987	0.295016	0.178378	0.761315	0.060307	0.980528	0.019472	[i, been, known, to, miss, a, red, flag, i, be...	[known, miss, red, flag, known, put, lover, pe...	[know, miss, red, flag, know, put, lover, pede...

50 rows × 39 columns

---