Introduction

The idea for Mood Music Matcher arose from my personal experience of searching for the perfect song to match my mood. I wanted to develop a tool that not only understands emotions but also curates' music to support mental well-being.

Solution

Mood Music Matcher addresses this by using advanced natural language processing (NLP) to analyze the sentiment of journal entries. It then suggests Spotify playlists tailored to the user's emotional needs.

• GenAI Model: OpenAI API

• Music Data Integration: Spotify API

• Orchestration Framework: LangChain

• App Framework: Streamlit

Github repository :mood-music-matcher

Deployed link :mood-music-matcher-maulcenter.streamlit.app

How to Use

1. Enter Your Journal Entry: Share your day or thoughts in the text area.

   

2. Analyze Mood: The app uses AI to determine your mood.

   

3. Get Music Suggestions: Discover songs that resonate with your mood.

   

4. Listen and Enjoy: Click the link to be automatically directed to Spotify and play the song.

   

   

Prerequisites

Set up a Spotify Developer Account

• Go to the Spotify Developer Dashboard.

• Create an application.

• Collect your Client ID and Client Secret.

Set up a OpenAI Account

• Go to the OpenAI website.

• Sign up for an account.

• Generate your API key in the dashboard.

Set up Environment Variables

In your development environment, set these as environment variables:

export CLIENT_ID='your_spotify_client_id'
export CLIENT_SECRET='your_spotify_client_secret'
export OPENAI_API_KEY='your_openai_api_key'

Code Overview

Here’s a brief look at the core functionalities:

Analyze Mood

The app utilizes OpenAI to determine the mood from your journal entry.

def analyze_mood(input_text):
    llm = OpenAI(temperature=0.7, openai_api_key=openai_api_key)
    mood_analysis = llm(f"Identify the mood of this text for music recommendation: {input_text}")
    return mood_analysis.strip().lower()

Fetch Songs

It uses the Spotify API to fetch songs that match the identified mood.

def get_songs_for_mood(mood):
    sp = spotipy.Spotify(auth_manager=SpotifyClientCredentials(
        client_id=spotify_client_id,
        client_secret=spotify_client_secret))
    results = sp.search(q=mood, type='track', limit=5, market='US')
    return [(track['name'], track['artists'][0]['name'], track['external_urls']['spotify']) for track in results['tracks']['items']]

Explain Song Choice

OpenAI provides explanations for why each song fits your mood.

def explain_song_choice(mood, track_name, artist):
    llm = OpenAI(temperature=0.2, openai_api_key=openai_api_key)
    explanation = llm(f"Explain why the song '{track_name}' by {artist} matches the mood '{mood}' based on its musical style and mood, without referring to the lyrics.")
    return explanation

User Interface

The app interface is built with Streamlit, allowing users to input their journal entries and receive song recommendations.

with st.form('journal_form'):
    journal_text = st.text_area(
        '✨ Describe Your Day, Today:',
        placeholder='E.g., Today was quite challenging, but I managed to push through and complete my tasks...',
        height=200
    )
    submitted = st.form_submit_button('Get Songs 🎶')

    if submitted:
        mood = analyze_mood(journal_text)
        st.subheader("Your Mood 🎭")
        st.write(f"It seems like you're feeling: **{mood}**")
        song_list = get_songs_for_mood(mood)
        st.subheader("Suggested Songs 🎵")
        for idx, (name, artist, url) in enumerate(song_list, start=1):
            explanation = explain_song_choice(mood, name, artist)
            st.write(f"{idx}. **{name}** by {artist} - [Listen on Spotify]({url}) 🎧")
            st.write(f"This song fits your mood because: {explanation}")

Conclusion

Mood Music Matcher is a unique way to connect with your emotions through music. By integrating AI and Spotify, it offers a seamless experience that makes your day brighter. Try it out and enjoy a personalized musical journey!

---

Feel free to customize further or ask if you need any more changes!
Special thanks to@HASHNODE

#AIForTomorrow #OpenAI #SpotifyAPI #Spotify