With the help of DPI identification and big data correlation analysis technologies,
HAOHAN DATA’S HH-TAG User Behavior Labeling System can look deep into users’ consumption records of music and videos, produce user behavior labels for refined use behavior and traffic, and analysis with a higher conversion rate.
²Collecting users’ click and consumption records of music and videos.
²Linked to music and video libraries, examining and tagging users’ music and video behaviors.
²Clustering users for inpidual user profiles.
²Allowing the queries of use labels, music records and video records.
²Providing user label data support to the third-party system via the interface server.
The system consists of four modules: the CDR screening module，music behavior analysis module, music analysis application display module, and label data interface module. Together they form the mobile user music labeling system.
²GDR screening module: screening DPI CDRs，and extracting information about the users' music behaviors.
²Behavior analysis modulo: linking to label libraries, clustering users' music behaviors, and generating music labels.
²Application display module: providing web-based data query and export functions including record query, label query, music/artist statistics, label statistics, etc.
lData interface module: providing interface to provide user tag data externally.
lProviding user label data support to the third-party system via the interface server.
Web-based data query and export functions which including record query, label query，music/artist statistics, label statistics, etc.
ØMusic record query
The system keeps records of the user’s music playlist and download behaviors, including the MSISDN, IMSI, TIME, song name, name of the artist, IP, RAT, cell-ID, and website, which are open to query and export.
Query and export of the user list by music, artist, any other label popularity.
Query and export of label statistical data by date or website:
• The total numbe of users (million people per day);
• The count of music plays(million times);
• Number of users and labels that are tagged in large categories dimension (cumulative by subclass);
• Number of users and labels that are tagged in subclass;
Query and export 〇f statistical data about the music and artists by time span or website:
• Music & artist: number of users and number of plays;
• Artist: name of artist, number of users and number of plays;
l Data interface
Data query interface for the third-party system Queries supported:
• Inpidual or combination of date, label, music and artist;
• MSISDN or IMSI queries;
Results returned: query conditions，label name, MISIDN and IMSI.