With the recent surge in interactive social applications on an internet-wide scale, the role of gofai machine learning techniques for solving knowledge-rich applications seems to be on a sinking boat. The phenomenon of Swarm Intelligence where actual intelligent people are helping each other solve problems is starting to manifest in neat ways! Thanks to folksonomy services and the web 2.0 infrastructure! For example, I have a feeling that people will stop using the current look and feel search engines like Google to find good and reliable information.
People are better today at most common real world tasks compared to the machines. Actually, people have just been using these machines to facilitate them concentrate on the harder problems and let these machines take care of the easier tasks. With interactive workbench systems coming up that allow people to empower these systems, people productivity is improving tremendously! I have a feeling that the mature web will be an intelligent network with personal people portals that will satisfy all his/her content based needs.
The need for machines helping us manage our content based needs are emerging in a different fashion - the object of these systems are changing from the content to the actual people behind it. Where is machine learning headed to in such a case - an interesting technical insight comes from John Langford's post.
Welcome to the new world of Human Computing!
Showing posts with label human centered computing. Show all posts
Showing posts with label human centered computing. Show all posts
Wednesday, April 16, 2008
Thursday, December 13, 2007
Synergistic Learning
There is a sudden surge in applications that leverage the overall user experience by enabling the user with technologies to look at what others are doing with the system and use that additional knowledge in novel ways to let the user make more informed decisions. These enabling technology systems come under the purview of recent educational programs like Human Centered Computing at the intersection of AI, Algorithms, Databases, Computational Semantics, Humanities, Psychology, Sociology and even Economics!
The goals from the point of view of creating better intelligent systems still remain the same, i.e., to do more intelligent machine learning, but now, by exploiting this Social Semantic Infrastructure by way of creating, adapting and learning from it and then studying the implications arising out of it. This is the high level knowledge intensive functional approach of looking at the problem of creating better intelligent systems.
This recent paradigm shift in computing of focusing more towards the 'human' aspect is a major leap for the not so old field of computer science which now, instead of being called a spin-off of mathematics and physical sciences is becoming a true mature science on its own.
Getting back to talking about the recent surge in semantic applications, these basically try to create systems that focus directly on actual task that the system is designed for by helping the user in simple but powerful ways. For example, collaborative and recommendations systems draw analogies from the 'wisdom of the crowds' idea to make decisions based on what the knowledge authorities have to say about it. Such fine-grained systems give the normal user the power to rely on the knowledgeable teacher of teachers that is helping the user carry out her tasks much more efficiently. This 'wisdom of the crowds' concept compares to the machine based equivalent as in statistical machine learning systems which draws inferences using alot of training data.
The emergence of so many novel social applications have been possible because of the recent semantic web standards for interoperability and the open source community projects that enable small teams to build large applications within a short period of time.
The goals from the point of view of creating better intelligent systems still remain the same, i.e., to do more intelligent machine learning, but now, by exploiting this Social Semantic Infrastructure by way of creating, adapting and learning from it and then studying the implications arising out of it. This is the high level knowledge intensive functional approach of looking at the problem of creating better intelligent systems.
This recent paradigm shift in computing of focusing more towards the 'human' aspect is a major leap for the not so old field of computer science which now, instead of being called a spin-off of mathematics and physical sciences is becoming a true mature science on its own.
Getting back to talking about the recent surge in semantic applications, these basically try to create systems that focus directly on actual task that the system is designed for by helping the user in simple but powerful ways. For example, collaborative and recommendations systems draw analogies from the 'wisdom of the crowds' idea to make decisions based on what the knowledge authorities have to say about it. Such fine-grained systems give the normal user the power to rely on the knowledgeable teacher of teachers that is helping the user carry out her tasks much more efficiently. This 'wisdom of the crowds' concept compares to the machine based equivalent as in statistical machine learning systems which draws inferences using alot of training data.
The emergence of so many novel social applications have been possible because of the recent semantic web standards for interoperability and the open source community projects that enable small teams to build large applications within a short period of time.
Subscribe to:
Posts (Atom)