NetEase Cloud music playlist analysis system based on Python data visualization

Table of contents

NetEase Cloud music playlist analysis system based on Python data visualization

If you have any study questions, please add me on WeChat to communicate! bmt1014
1. Project Introduction
(1) Project Background 1
(2) Project Process 1
2. Project design flow chart 3
(1) Overall architecture of NetEase Cloud music playlist analysis system based on Python data visualization 3
(2) Obtain information on the playlist index page 4
(3) Get information on the playlist details page 5
(4) Top 10 number of song appearances 6< a i=9> (5) Top 10 European and American playlists of NetEase Cloud Music 6 (6) Top 10 European and American playlist reviews of NetEase Cloud Music 7 (7) European and American Distribution of the number of playlists played 7 (8) NetEase Cloud Music European and American playlist label chart 8 (9) Songlist introduction word cloud chart 8 3. Project implementation code (the code is too long, not included) (1) netease_cloud_music_data_analysis.py 9 (2) music_index.py 11 (3) music_detail.py 13 (4) top_10_song.py 15 (5) top_10_ea_song_playlists.py 17 a> (11) music_wordcloud.py 28 (10) label_ea_song.py 26 (9) top_10_ea_song_playlists_distribution.py 25 (8) top_10_ea_song_collection_distribution.py 23 (7) top_10_of_ea_song_comment.py 21 (6) top_10_of_ea_song_collection.py 19 (1) Top 10 number of song appearances 31 (2) Top 10 European and American playlists of NetEase Cloud Music 32 (8) The commercial value of song list data visualization 37 (7) Music playlist operation analysis 35 (6) Songlist introduction word cloud chart 35 (5 ) NetEase Cloud Music European and American playlist label chart 34 (4) Distribution of the number of European and American playlists 33 (3) Top 10 European and American playlist reviews of NetEase Cloud Music 32

1. Project Introduction

(1) Project background

With the popularity of music software, massive amounts of related data have been created. In the era of big data, any large amount of data will generate huge value once it is utilized. There are many examples of using Python to analyze song-related data to discover customer needs and further expand the number of users.
Taking into account the practical operability and Python has very mature libraries in data analysis and interaction, exploratory computing and data visualization. After the team tested the feasibility, they decided to use Python to analyze the music software playlist.

(2) Project process

This project uses Python to obtain NetEase Cloud Music playlist data and conduct visual analysis of the playlist data. Get the quantity graph and word cloud of the song list’s comments, collections, plays, contributions, and distribution, and make suggestions for song list optimization.
The project uses crawlers to obtain data, then cleans the data, and finally performs data visualization. During the analysis process, numpy, pandas, matplotlib, time, requests, squarify, jieba, wordcloud, and bs4 third-party modules were used. Finally, bar charts, word cloud charts, and tag charts were used to display song collections, playback volume, and other related analysis results and combine them. Relevant data optimizes playlist playback volume.
Finally we implemented the project and tested the project.

2. Project design flow chart

(1) The overall architecture of the NetEase Cloud music playlist analysis system based on Python data visualization

(2) Obtain information from the playlist index page

(3) Obtain information from the playlist details page

(4) Number of times the song appears TOP10

(5) NetEase Cloud Music European and American playlist TOP10

(6) NetEase Cloud Music European and American playlist reviews TOP10

Figure 8 Flowchart of NetEase Cloud Music European and American playlist review TOP10

(7) Distribution of playlist numbers in Europe and the United States

Figure 9 Flowchart of the distribution of playlist numbers in Europe and the United States

(8) NetEase Cloud Music European and American playlist label map

Figure 10 Flow chart of NetEase Cloud Music’s European and American playlist tag map

(9) Song list introduction word cloud chart

Figure 11 Flowchart of song list introduction word cloud chart

3. Project implementation code (too long, omitted)

4. Project analysis results

NetEase Cloud users are a user group who are very accustomed to creating playlists, and many users also learn about some songs through playlists. A NetEase Cloud user can even create hundreds of playlists by himself. We crawled data from NetEase Cloud's European and American playlists and created the following table.

(1) Number of times the song appears TOP10

Figure 12 NetEase Cloud Music European and American playlist TOP10 songs
We have obtained the TOP10 songs in NetEase Cloud Music European and American playlist. I believe that friends who are familiar with European and American songs will be familiar with these songs. Very familiar, they may have appeared in the playlist recommended to you before.

(2) NetEase Cloud Music European and American playlist TOP10

Figure 13 Top 10 European and American playlists of NetEase Cloud Music
This is the top ten list of European and American playlists. The first place has a play volume of up to 400 million. It can be seen that European and American songs is very welcome. From the statistics in the playlist, we can see that melody and rhythm are valued and liked by many European and American music lovers.

(3) NetEase Cloud Music European and American playlist reviews TOP10

Figure 14 NetEase Cloud Music European and American playlist reviews TOP10
You can see that the playlist [Customized, European and American recommendations that know you best, 35 songs updated daily] has the largest number of comments. This may be related to the nature of customization in the future. It can be seen that many popular playlists usually have many comments. The playlists in this list can also be found in the previous list of popular playlists.

(4) Distribution of playlist numbers in Europe and the United States

Figure 15 Distribution of the number of playlists in Europe and the United States
Logarithmic processing of the number of plays allows us to intuitively see the distribution of the number of playlists.
The number of playlists is mainly distributed between 0-10 million.
where ln(10000000) = 16.

(5) NetEase Cloud Music European and American playlist label map

Figure 16 NetEase Cloud Music European and American playlist tag map
We also found the tags of these playlists. Since it is a European and American playlist, [European and American] naturally accounts for the majority, almost half. We can also see that there are many categories of [Electronics] and [Popular]. [European and American Popular] and [European and American Electronics] have become more and more popular among NetEase Cloud users in the past two years, especially [European and American Electronics]. Many Young people are already big fans of electronic music.

(6) Song list introduction word cloud chart

Figure 17 Song list introduction word cloud diagram
Song list introduction word cloud diagram. These words include the style, background and story of a song list. Maybe you have collected this song list because of a certain word that moved you. Resonate with the song.

(7) Analysis of music playlist operations

There are two forms of playlists, one is the playlist created by the official, and the other is the playlist created independently by users. In NetEase Cloud, most of the songs we collect and listen to come from individual users. The operation of playlists is meaningful to both individuals and officials.
Many users use playlists as an entry point to learn about music, and rarely actively search for music continuously. Therefore, whether the playlist is rich and whether the quality of the playlist is good enough determines the user Frequency and duration of use. The more good playlists users can get, the more valuable this music platform is.
For individuals, creating a playlist that satisfies many people allows the individual to gain attention and a higher sense of accomplishment, thereby creating a better quality playlist.
Therefore, based on the above analysis, we made an operational analysis of European and American playlists:
In terms of playlist content, Figure 15 NetEase Cloud Music European and American songs In the single-tag picture, we can see that popular tags and electronic tags account for about 50%. Implementing personalized recommendations based on tag classification can make playlists easier to find, expose, and manage.
In addition, our playlist can move closer to these two popular directions, targeting a wide audience and gaining considerable traffic. Coupled with the rise of electronic music, more and more excellent electronic music appears, which can enrich our playlist.
In terms of the title of the playlist, the title of the playlist is equivalent to an article. The title of the playlist can attract the attention of the audience and let them come in to learn about the playlist and listen to the playlist. Among the TOP10 playlists and the TOP10 playlists by number of comments, we distinguish the following categories:
In terms of special symbols, we can see " 【】","|" and other special characters can attract the audience's attention and emphasize the highlights of your playlist.
In terms of strong attraction, use words such as "selected", "10w+", "exclusive", etc. to drive the audience to click.
In the conclusion, words such as "Blood-burning | Refreshing and anti-drowsy BGM" are used to drive the audience to click.
In summary, the success of a playlist is inseparable from the title, song style, and even the number of songs. Only by balancing these factors can its playback volume reach expectations.

(8) The commercial value of song list data visualization

Visualizing playlist data can help music companies better understand consumer preferences and provide consumers with playlists that better suit their preferences.
In addition, playlist data visualization can help music companies better analyze the information in playlists and use this information to decide how to price, market and distribute music.
Therefore, the visualization of playlist data has certain commercial value and can provide important market information for music companies.