Fake personnas and the variables needed to screen them...

While reading a thread, I may have found some similarities between some posts that would indicate a single source. The "source" would not necessarily mean a single person, but a group, pushing the same ideas or actions in the shortest time possible.

If an unwanted topic from these groups appears, all they would have to do is flood the said topic with their similar rhetoric in the first pages. Now, in order for a software to do that automatically, a database of synonyms would be required at least. Knowing that and the fact that an idea can be expressed in many but limited ways, it would be theoretically feasable to make a software to counter-act this kind of forum manipulation.

If I thought about it, "they" surely did. So, again with the word "flooding", but inside the posts themselves, I thought. And for that, that would be a "mid-text neutral comment" or "useful quotations" database. The kind often taken out of their context to convince people that won't or don't have/take time to read.

The last and the most important variable in this equation is you, the reader/poster. Acting quickly when in direct danger is primordial. When you enter a forum, you are not in a fight or flight situation. Take your time to think about the argument you're about to post, everyone will benefit from it, including yourself. This advice I gave you, I try to follow it myself. But the "fibrofog" isn't helping.

These are only the basic variables. There should be many more to find.

reply to post by Vio1ion


From logging a number of forums at political websites recently there certainly is enough of a pattern in certain logical associations and responses to prove beyond a doubt a common origin I think you are looking at a chat-bot/search engine/human operation similar to how time-life use to run there marketing operations. My guess is that a very large number of topics across a broad spectrum of websites can be "washed" by a relatively small group (perhaps five individuals) but the underlying logic betrays a common source for the injected material .

Due to the nature of the operation, avoid getting sucked into short "If , than" scenarios as they seem good at locking on and cognitively interacting (and responding with "get off the path" hooks)with those types of posts . good solid thinking and tying highly contextually sensitive words seems to limit the ability to respond signifigantly

I know what you mean, generally the first page has most of the debunkers trying to quickly shoot down the thread when it first appears. After some time and get a bit deeper into it more reasoned minds have some time to contemplate and express more thoughtful messages. There are still some topics like the 9/11 threads where a conceded effort is placed to push the official story, but the lengthy discussions are more individuals behind it rather than bots.

As for trying to shot down these bots, excellent work in bringing up the topic. As a programmer the thought of natural language processing really does my head in, it will take many years to master as each word is a variable with its own associations and things get complex quickly. Eliza is one of the first programs making steps into natural language and heaps about it on the net. If I was to take this on I would be looking towards a neural network programming structure to parse the message and derive a conclusion.

I could see a point to this logic, but it would be difficult to differentiate between actual groups of people who are deciding to flood using authentic individual accounts and a flood of bots.

As can be seen by the U.S. Republican party, many individuals can switch to a new rhetoric paradigm very quickly and begin flooding a message utilizing the exact same wording as the others. These are all individual people, but if you were to type out their statements it would look like the words of one person.

I am not saying that that sort of tactic makes for good discussion, but it is technically a valid tactic.

If one poster is suspected to be fake, we'd have to look at all of its posts to spot repetitive patterns. We could even start threads with about the same subject, each worded differently as a kind of trap, to try to find the precise ideas targeted by the "fakes".

Another variable that came to mind is the way a thread gets derailed, as there should be a pattern of teamwork if "fakes" are used that would give the illusion of a majority. This illusion is quite important because most people prefers to follow than lead. And that brings up the psychological variables. Again, a pattern should appear because it is quite unlikely that two people with the same psychological profile would share the exact same ideals.

Then, there is the matter of personnal anecdotes. We already know that this kind of software will provide a background. But does it provide more specific details, like childhood memories ? Everyday life anecdotes ?

In my experience there often is a "source" for a slough of very similar threads... Political commentary shows, as well as a few other websites and webblogs.

If one pays attention, one can often find individual interpretations (in some cases regurgitation) of whatever the talking heads have said in the last 12-24 hours, all from several different posters. For example is Joe Shmoe goes on his talkshow tonight and says "Congressman Smith is a psychopath, and here's why...." often, and within hours, the barrage of "ZOMG SMITH IS A PSYCHOPATH!!!!" posts and threads hit directly afterward.

This is especially true in the political forums.

But this is not an indication of any plot or grouping. It's simply a group of like minded people using the same material as source and trusting the same journalists/bloggers to do their footwork.

~Heff

reply to post by Hefficide


Good points, but I'm sure that if fakes are used, it won't be as simple as "ZOMG" + adjective + "not so credible source". People like me will just skip these too simple replies. The goal is to manipulate in a certain direction, and it would'nt work with only an opinion.

»»Now I'm wondering...

Could the chinese use such a software ? Or more simply use a fraction of their population to overwhelm the net, hidden behind proxy servers. Also applicable to other nations with a large population and more frightingly, the possibility of a combined effort united by common goals from 2 or more countries.

»»End of wondering

While it's in my head, I'd like to mention that things are rarely completely black or white. In between, there is what I call multiple shades of grey. Each facts should be put in their context and analysed logically. But maybe too many times, one's sense of morality or ideals will get in the way.

Originally posted by Hefficide
In my experience there often is a "source" for a slough of very similar threads... Political commentary shows, as well as a few other websites and webblogs.

If one pays attention, one can often find individual interpretations (in some cases regurgitation) of whatever the talking heads have said in the last 12-24 hours, all from several different posters. For example is Joe Shmoe goes on his talkshow tonight and says "Congressman Smith is a psychopath, and here's why...." often, and within hours, the barrage of "ZOMG SMITH IS A PSYCHOPATH!!!!" posts and threads hit directly afterward.

I have noticed a similar effect when looking at the stats of my website. For instance, I have a story about the New Madrid fault line I published years ago, and now, because of the media attention, I get more hits for this. You can have top ranking in the search engines for any number of topics, but unless it gets media attention, hardly anyone searches for it. People see it on TV then look it up and makes comments in blogs, etc.

reply to post by Hefficide


Certainly, it is easy enough to find the meme points from the talking heads and then identify that as a common source (as media, functionally speaking, is a simple social engineering tool), what creates the deviation from a bell curve (the lovely brownian motion of groups of minds at work) is the application of certain systems of logic.

That is what glows in the dark. I mean regurgitators generally can't carry the logic of certain propaganda because they haven't been 'given' the answer to objections from the common source. So that discrepancy can be use to draw out an expert system (and it's operators).

Practically it breaks down to following both the number and nature of posts from a given individual (or group of ) and how or when they actually respond to changes in logical paradigms. The beauty of the system is real people posting same sourced info and helping cloud the algorithms for search and auto insertion.

An interesting experiment is to insert known real responses into existing good chat bots and evaluating how few changes would have to be made at a data/library level to get a very broad response culminating in a collective tone.