A message bot is a computer program that facilitates chat or actions on messaging platforms like Twitter, Facebook, MySpace, etc. Message bots go by several other names as well. They are sometimes referred to as “web bots”, “web conversation bots”, or simply “robots”. The idea is basically the same though.
In a network environment, messages sent are generally interpreted by all nodes in the network. nodes are computer programs which are connected via the Internet and/or other connection technologies. They can be software agents or real people. In some cases, these “robot” or “combot” connections can be subdivided into smaller groups of communication nodes or “neighborhoods”.
For example, if you have a message client on Twitter, and you send a tweet about a specific event happening in your company’s office, it is forwarded from one bot to another bot on Twitter, whose owner you have never met, and one from an individual in your company’s office who is not authorized to tweet about such an event. These “boutiques” or “connections” are a lot like a chain reaction. The first bot notices the tweet, gets on their computer, uses their “robots” to forward the message on to additional “robot operators” who in turn receive the message from the original bot. In a very short time frame, several different conversations begin happening all over Twitter.
Of course, there are also more sophisticated methods to accomplish this. Most bot conversations utilize message-hooks, which allow a series of commands to be passed or re-requested during each conversation thread. For instance, in some Twitter examples a bot may allow a user to post a command like; “ping @garystuplite”. When that command is passed along to other bot operators, those bot operators will “ping” back using their own “robots”. This allows multiple conversations to happen without much human intervention.
There is another interesting aspect of these conversations. Each of the users, or their respective bots, are defined by two long conversations. The first conversation creates the argument description for the message. This argument describes what the user is doing at the moment, what their intent is, or the message they are trying to convey. The second, and final conversation create the reference document for the bot.
Reference documents are essentially logs of the conversations that take place. Each log has a reference page, which can be accessed by clicking on the conversation threads. This page contains the name of the user who started the thread, the date and time they began the conversation, and a number identifying the bot used for the conversation. In addition, each log contains a “meta-data” description which gives further information about who was logged in, how long the conversation was, and any other useful data. Meta-data are especially useful when identifying multiple users for testing purposes.
The message bots can also be taught how to perform action functions such as making reply messages, inserting link objects, and posting comments. Bots can also define and set rules for performing each action. Bots can use the built in reply message system to create a message, save it as a message, and then send it out as a reply to a message thread. The bot can also define and set rules for inserting a link to another website within a reply message, and then send out that message along with an optional body copy.
The event types that can be sent with a Bot include: generic text, image and audio, game and application icons, and web browsing. Generic text is just a message that can be recognized by most text-based applications and websites. Image and audio are of the “audio only” variety and cannot be recognized by search engines. Game and application icons are those that make sound or show visual stimuli. Web browsing allows the user to visit a site and view a particular page.