Use Case - Social Network
We are glad to have among our partners a social networking service. The service includes a website, Internet messaging service, and other features like posting and viewing pictures and blogs. It allows its members to create unique personal profiles online in order to find and communicate with old and new friends.
The services are for the personal use of members only and may not be used in connection with any commercial endeavors. Commercial advertisements, affiliate links, and other forms of solicitation may be removed from member profiles without notice and may result in the termination of membership privileges. Despite this, there have been a lot of violations when users register just to illegally post their commercial information. This makes the site less attractive for its members and visitors.
Before
Now the network has more than 20, 000 members. Every day a few more people become registered and create their personal profiles and blogs. Each text has to be approved before being posted. Automatic verifying with traditional statistical tools did not satisfy the administration. The company had to employ more editors to test new profiles and blogs for illegal advertising.
We can't just delete a message which has the words sale or price in it. More than that -- the spammers are very creative. They avoid the words which are usually used as indicators. We want to keep the main feature of our service: an individual approach to each member. But it takes too much time and effort.
An idea popped into my head when I saw the ad at a software forum. It said,
Summarizer will read Web pages for you!
At that moment it was my burning desire to find something that would read those endless blogs for me! Surely it does not do my entire job, but it does its boring part and does it fast!
Now
The corporate system developers have made use of a function that allows creating theme-oriented summaries. Summarizer Pro generates "Economics" summaries where the main theme is economy and commerce.
After a member has posted their message, Summarizer starts doing its job. It generates a summary of the text and forwards it to the editor. If the editor sees that there is no commercial information in the summary, the message is posted right away. Otherwise, they return to the text and make their decision about posting or banning it. The next step in site development is to create a page with recently-posted blog reviews. These reviews will be automatically generated and grouped into catalogues by Summarizer.
This page will be available for members and visitors. Reviews of the blogs will be sorted by date and main topic. There will also be a link to the author's profile. We think this feature will push up traffic within the network and attract more users.