Social media is the next evolution in how we use the web. However, with all the great things that have come from our new found ability to connect on the web, we’ve also made the creation of spam much easier.
Nowadays, it’s easy for almost anyone to create a presence on the web. With our quest for ease, we’ve made it easy to automate the process. Until we’re able to differentiate spam from actual information, we’ll be unable to properly index and search the social web.
Problems:
One of the biggest problems with social media is that currently, many websites allow you to create unlimited profiles. One person can create hundreds of Twitter accounts, each of which can send out a Twitter message, quickly cluttering the web.
A second problem is that unlike the current web page structure, conversations can jump incoherently. Consider a conversation taking place on Facebook through status messages. Each individual status update would make little sense, however collectively they can offer a large amount of information.
Also, we currently have little ability to to understand true connections and consumer intent. What I mean by that is, how do you differentiate between a spammer friending a spammer versus a true friend? Profiles can artificially inflate connection numbers, making them seem more important than they actually are.
Lastly, the overall speed at which new data is introduced creates an issue for search and indexing. Right now, we’re unable to quickly assign authority and trust to new content. We can assign trust to specific users, but lose out on new influential users.
Examples of Problems:
The most powerful example of the problems with the social web is Twitter. Consumers can quickly and easily create multiple profiles, and send many messages.
Not only are spammers able to disrupt the streams of other users, through @ replies and direct messages, but each individual profile and tweet creates a page to be indexed in our conventional search engines.
Lastly, each tweet also influences the latest trend. With enough profiles and messages, a spammer can manipulate what topics are trending upwards. On the social web, topics trending offers insight into the latest news. Obviously, being able to affect this nefariously, can quickly make trending data useless.
What Needs to Change:
The first things that needs to change is our ability to detect spam signals. Determining patterns in how actual humans interact with each other on the web may give use ways to differentiate between an actual user and spammer.
Secondly, introducing an API or data sharing system to find patterns might be needed. The data need be anonymous, because of the privacy issues associated with it. Without all the information, our understanding of the social web will continue to be incomplete.
Lastly, we need to change how we return search results. The current system of ranking only by importance needs to evolve to include temporal affects. A trustworthy result should not always outrank a semi-trustworthy result, if the latter is newer. Newer pages should be given a greater weight for rank.
What is Happening Now?
Right now the closest thing we have to a search engine for the social web is Twitter (scares the crap out of Google). However, this “search” does not separate out spam from trust. Neither does Twitter search assign authority to influential
Patents:
Method and Apparatus for Detecting Spam User Created Content
The present invention provides methods, apparatuses and systems directed to automatically detecting spam user created content. In a particular implementation, there is provided a method for processing spam contents, which comprises: maintaining a plurality of key information databases; receiving user-created content and at least one of a service ID and a content category ID of the user-created content from one or more users of a user-created content hosting site; selecting one of the plurality of key information databases based on at least one of the service ID and the content category ID…
System and method for developing and using trusted policy based on a social model
A trust policy is constructed based upon a social relationship between real-world entities. The trust policy may determined based upon a social network and social network maps. The social network map provides a framework to determine social distances. The trust policy provides quick and secure access to desired or trusted nodes while providing security from entities outside the trusted sphere of nodes. The trust policy determined by the social distance may be used for various types of applications including filtering unwanted e-mail, providing secure access to resources, and accessing protected services…
Social Analytics System and Method for Analyzing Conversations in Social Media
Conversations in an online content universe are monitored. A social analysis module analyzes individual conversations between publishers in the online content universe. Publishers that influence a conversation are identified.
Related Videos:
Semi-Supervised Learning: A Comparative Study for Web Spam and Telephone User Churn by David Siklosi
IR in Social Media (IRSM) by Matthew Hurst and Alexey Maykov
Check out Training Social, a comprehensive resource that will help you build and execute a social media plan for your business!
{ 6 comments read them below or add one }
It’s almost as if we have to be two steps ahead of spammers. They continue to adapt their methods to changes in technology.
The spam on twitter does suck. So many people with no intentions what so ever. But at the same time many of these people simply dont know how to use the system. A lot of new users get on friend a million people and start promoting their stuff cause thats what they were told. They dont always mean to do these things just they dont get the entire system.
Very tough to sort out.
Very hard to differentiate between spam and real twitter messages. With the introduction of intelligent spam detection comes the inevitable introduction of intelligent spamming. Its all part of the life-cycle of many web technologies.
It’s getting out of control Joe, hopefully things change soon.
It’s sad that spamming has become more than bothersome these days.