CIC “Blog Content Mining & Analysis” at CNBloggerCon 2008 Guangzhou

China social media and the world: US 220M Netizens; CN 253M; 74M bloggers in US, over 100M in China.  Blogs and how they reacted to the Wenchuan earthquake. Prayers shown. In wake of quake, blogs became a platform to express feelings.

Commercial value of blogs: competitive intelligence, product feedback, effectiveness measurement, reputation monitoring, digital PR, more.

Commercial value: Marketing inspiration. LISTEN, KNOW, PARTICIPATE.

Of note: How Chinese language is broken down.

ZSS: CIC prezo close to life, easy to follow; must blogs make money?

How words mix and match in relatively different phrases shown. It’s complex and a bit like Lego + mixing and matching.  Also: how keywords and grammar mix and match; analysis based on this.  Different types of analysis: keywords and word-structure-based methods. Twitter now shown in screen to right.

http://jiong.ws — the character 囧 (which basically shows an upset face) is VERY big in Sinosphere Web (CN, HK, TW). Also mentions 火星文 (TW).

Content category: industry/brands/products; how people feel; how users express their feelings.

As the talk drags on, the ZSS is full of people talking about favorite cars. We smell Porsche, BMW; the DF favorite is a BMW 500-series.

Data needs to be aggregated, analyzed in different angles. Q: How many articles talk about each brand per month; more.

More Q: How many people talk about each product per week; what’s the top 10 people who are on about a particular brand?

Solution: OLAP data cube (fast/flexible data analysis support). Interesting 3D graph used.  How to make more sense of the data: a web graph shown. How the Top blogs link to each other: Top 1 in center, woven with other Tops.

Very sharp question about @isaac:

Q: How does your company relate to bloggers?

A: Everyone wants to keep free speech; also innovate at same time.

Questions are now taken from the floor.

Q:  how can you make sure the stats are real?

Netease livecast now audio-only, it seems.  Lunch is available shortly. Announcements are written all in standard traditional Chinese. See you soon

More info at: http://blog.csdn.net/cictech