Weibo \’s gathering place as a hot topic, evaluation, like-it-and-greet forwarding, and commentary with a waterless army…may all affect new people\’s judgments. This article will take a look at the user behavior data on the marriage policy microblogging. Let\’s take a look at the mysteries of microblogging user behavior that are worth digging!

                                                   ▼

Every year before and after the “May 1st” is the get-together of new people\’s marriage. With the advent of the wedding season, how to do a wedding that is unforgettable for life has become a hot topic. Weibo\’s gathering place as a hot topic, evaluation, like-it-and-greet forwarding, and commentary with a waterless army…may all affect new people\’s judgments.

In view of this, tecdat researchers conducted inventory of user behavior data on the marriage policy Weibo, combined with semantic analysis, positioning the user\’s interaction, in order to extract valuable information.

Interactive Analysis: Users are most concerned about wedding brands and cases

In this analysis, we collected a total of 13,585 Weibo, 678,295 keywords, as well as corresponding favorite, forward, and commentary for each Weibo.

And configure the weights: forward 0.5, comment 0.3, like 0.2. Calculate the weighted value (recorded as an interactive score). Based on: The weighting algorithm used by the company\’s internal applications.

Display

In order to facilitate the measurement of interaction, we define that: the interaction score is less than or equal to the lowest 25%. The interaction level is recorded as low, and the 50%-75% interaction score is recorded as high, >= 75%. Interaction points: The interaction level is marked as max (excellent).

Display

The most active vocabulary is “wedding”

The highly active vocabulary is closely related to the company\’s name: a total of 4967 microblogging marketing a good number of years of wedding, a hundred years there have been 1090 thousand lilies a total of 2718 microblogging, Qian Li appeared 1090 times silver wedding planning company a total of 1754 Weibo, only the high-interaction keywords in the silver bullion have appeared 516 times, the above data is only the data distribution of high-interactive Weibo.

Keywords related to microblogging wedding planning case, high user activity (ie: more effective users of non-zombie powder)

Scene subdivision: content, description, color

Then we choose highly active words about event planning to analyze word frequency distribution. Select planning related keywords for interaction level max:

“Wedding” interactive score 3717660, accounting for 5.07%, topped the list;

“Forwarding” interaction score 2799078, accounting for 3.82%;

The “hotel” interaction accounted for 2.71%.

Wedding accounts are promoted on Weibo, and “rewards” and “drafts” activities are frequently held around the theme of “wedding .” According to Weibo active points data. This approach works best in terms of comprehensive forwarding, comments, and likes.

Helping customers intuitively understand the topic of microblog posts, using forwarded draws, triggering potential customers to be active and eye-catching, and helping to increase the exposure of microblog accounts.

Display

Select content attributes related to the interaction level max

The Weibo content mainly includes case pictures, mainly introducing weddings and hosting “hotels” , weddings involving “brands” and wedding “photographing cameras”, etc., mostly based on pictures.

Select relevant keywords for the case with interaction level max:

Users prefer the case with advanced attributes. The color of red , blue, and gold is most liked by users.


▍关注我们

大数据部落 -中国专业的第三方数据服务提供商,提供定制化的一站式数据挖掘和
统计分析和数据挖掘咨询服务:y0.cn/teradat(咨询服务请联系官网客服

点击这里给我发消息QQ: 3025393450


【服务场景】        

 

             

科研项目;

        

             

公司项目外包;线上线下一对一培训;学术研究。
【大数据部落】提供定制化的一站式数据挖掘和统计分析咨询服务

 
分享最新的大数据资讯,每天学习一点数据分析,让我们一起做有态度的数据人【大数据部落】大数据部落提供定制化的一站式数据挖掘和统计分析咨询服务
微信客服号:lico_9e
QQ交流群:186388004  
欢迎关注微信公众号,了解更多数据干货资讯!

 

版权声明:本文为匿名原创文章,遵循 CC 4.0 BY-SA 版权协议,转载请附上原文出处链接和本声明。
本文链接: