RunningTalks platform enables people to share their opinions about talk subjects created by people themselves. People record and post their comments in audio on existing running talks via tools provided by the RunningTalks platform itself.

While everybody can browse running talks and listen to others' opinion, but only registered users can post their comments on existing running talks and create their own running talks to ask people's opinion about anything they can come up with.

Components of RunningTalks Platform

RunningTalks platform consists of a website www.runningtalks.com and a software application, called RTClient. In order to browse running talks, create new ones and post your comments, you can use both the website and the client since they provide same functions.

Our software application RTClient is used for accessing RunningTalks platform remotely from your desktop, so that you don't have to use any web browser. You can also use RTClient without having to sign up. It also provides some customization settings. Unfortunately, RTClient runs only on Windows machines.

Goals

We believe, RunningTalks platform has a very important role with ambitious goals to accomplish. Today social networks are the most convenient grounds for socialization and collaboration among people on the Internet and combining free speech act with the social networking concept is yet another untaken step on the Internet.

While embarking on that road, we have these goals in our minds:

Helping people share their experience

Being a platform for free speech

Providing innovative and fast user experience

Being a testbed for our future web-based applications

As we keep this platform up and running and more and more people start to make use of it, we will determine more specific goals for better and faster user experience.

Support RunningTalks Platform

We believe RunningTalks Platform will attrack many people from all countries around the world. Everybody will try to make use of the platform as much as possible by posting their comments and asking peoples' opinions. More people means more different opinions to listen to. We hope every member of this community help us spread the word so that everybody can get chance to say something about things happening around us.

To support RT, we recommend you to share your own running talks with your friends, readers and listeners and let them comment on your talks. RT is especially good platform for brainstorming, so inform as much people as possible about your talks and let them brainstorm.

Currently, RunningTalks platform is funded by donations of its community members and revenue of other projects developed at the Ziemantics Labs. We strongly recommend you to donate or purchase our products Ziepod+ or Zietag and help us afford the expenses of our servers and the bandwidth. Your donations will come back to you as new features and fast access.

About Ziemantics LLC

Ziemantics LLC was founded in the August of 2007 in Delaware by Arda Celebi, who is a computer engineer and the lead developer of all projects being developed at Ziemantics Labs.

Ziemantics LLC is developing tools to process, manage and share information. RunningTalks platform is the fifth project developed at the Ziemantics Labs and the only one that is concentrated on information sharing concept.

Other than RunningTalks platform, Ziemantics Labs also hosts Ziepod, Feedzie, Newzie and Zietag projects. You can learn more about these projects and the vision of Ziemantics LLC at www.ziemantics.com.

All in all

Please visit www.runningtalks.com and get into the platform. We hope you enjoy your time there and express yourself freely. RT is at the BETA stage so please report any anomaly you can find thru our feedback panel. And don't hesitate to add your suggestions as well. We are looking forward to hearing your thoughts about RT.

Have a nice day,

We believe being a part of a company means a lot. Like every company, Ziemantics has high goals to accomplish, in order to provide you the best. Feedzie will be part of that plan and will gain strength from every part of this new born company.

We like you to visit www.ziemantics.com to learn more about Ziemantics LLC and our goals. Hope you like the vision we have in mind and support our company by mentioning it to your friends, readers and listeners.

Have a great weekend,

So the task is simple; order podcasts based on how content centric they are. For example, in case of Technology category, those podcasts that mostly mention technology oriented words in their published content should be listed at the very begining of that list. Ok, here are the links for each major category and then you will decide how well our categorization works.

Most related podcasts for category 'Business'

Most related podcasts for category 'Entertainment'

Most related podcasts for category 'Life'

Most related podcasts for category 'Politics'

Most related podcasts for category 'Science'

Most related podcasts for category 'Sports'

Most related podcasts for category 'Technology'

We have tens of thousands of podcasts listed at Feedzie.com however each of these specially ordered lists contains no more than a couple of tousands podcasts. That is because we filtered out the ones that are not content centric, in other words the ones that publish content in variety of categories but not on specific one.

Hope you make use of this list to find new podcasts that you might like to listen to. And, of course, we recommend you our own podcast receiver for listening to podcasts, that is Ziepod.

Mapping Unknown to Known

Can we teach machines how to understand and interpret what any given story is telling about? Not exactly ... but we can ... to a certain extent. We can map the content to predefined concepts that can be interpreted by a machine and then let it interpret those concepts rather than actual words. It's a clever trick. If you don't know it, think of it as a different thing that you already know.

In case of Feedzie.com, we programmed our modules to learn which words represent which concepts and then make use of that data to categorize feeds and their published content. Mapping unknown to known cannot be done without loosing any information and the lost data depends on how we select those concepts onto which we map the actual content.

Selecting concepts, that is, categories

Every word has something to say but not all of them means something on its own. In other words, existance of some words in a story doesn't imply anything about what that story is all about. They somewhat connect other words that give the actual meaning to that story. Think of prepositions. Sounds like miscellaneous, doesn't it? So, this 8th category somewhat helps other 7 categories cover up the conceptual space without leaving any word outside.

If you want to define your own categories, you may not want to use that "miscellaneous category." That is perfectly normal. This is the way we map the real world into categories. Yours can be different. In fact, except the 8th one, we used 7 major categories but in real life, it is believed to be around 50-60 conceptual categories.

Task to accomplish : Learning

So where are we? Ok, we are trying to teach our machine to categorize given content and we just determined our categories. Now it is time to show our machine what is what and let it learn the relations between words and categories. To do that, we provided training data that consists of more than a million stories along with the data telling which story is in which category.

Since we select small set of categories, the real problem behind learning is not to choose how to design our learning strategy, it can be as simple as counting occurances. However, the real problem is finding large enough training data so that our training module can see all the words in high enough cases.

When we feed our training module this training data, it looks at which words occurs in which categories and extracts statistical relations out of that. It simply counts how many times it occurs in one category and then normalizes with the total occurance and finds the percentage. After finishing up training, it should be able to tell which words are most likely related to which categories, such as word 'Google' 70% of the times occured in 'Technolgy' category hence it is most likely related to 'Technolgy' category. Pretty simple.

After longs hours of traning and couple of naps(remember we are talking about millions of stories), our trained module knows how to map words to categories and find the dominant category by considering all the words of a given story.

Where are we at categorization?

So far what we told you seems like a mathematical calculations, and it is actually, simple as counting jumping sheeps. But the outcome is not that abstract because--after all--assigned categories are fairly good enough to describe the content and you can easily and effectively use them to optimize search results and to understand the content of a feed at one glance, like the way shown below.

So, we started with statistical models and ended up with content-oriented tool that you can make use of it to define what you are looking for in a better way. That is our job, to develop tools to let you define what you have in mind.

Failing at categorization

There can be various factors that fails us at categorization task. Those can start from the beginning by not selecting the right set of categories, or a result of a small set of training data. Even if your learning strategy is good and training data is fat, the story you have to categorize may not have enough words to determine its category. Or more dramatically, it can be in another language. Not fair huh? Welcome to the real world, my statistical friend.

Being better

Recall that we only defined 7 major categories. It is a good start but sure it can be improved. How? we can either increase that number or determine subcategories for each category, which help us focus on more specific content and let our training module learn more specific relations.

As we increase the number of dimensions onto which we map actual words, we should be able to preserve more information during that conversion, which leads to better understanding of relations between words and concepts. This also gives you more options at optimizing your search by allowing you to specify what you have in mind in more detailed ways.

In fact, if you think further, you can easily see that tagging is also another form of categorization where each tag corresponds to a category, which makes those calculations a real mess. But it is a subject of another post. Lets give you a break.

As this is the first post of Feedzie Blog, let us introduce ourselves by mentioning the history of Feedzie.com and our upgraded goals.

Short History of Feedzie.com

Feedzie.com made its debut in 2005 along side of Newzie, a news aggregator. Our first goal at that time was to let our visitors, especially Newzie users, find the feeds they might be interested in. Other than searching, Feedzie also provided tags to be used at browsing feeds.

As time passed, we released another software of ours, called Ziepod. As it is a podcast receiver, Feedzie became more focused on podcasts, especially audio and video episodes published thru those podcasts. From the first days of Ziepod till now, Ziepod users can search audio and video episodes from inside their Ziepod client thanks to the Feedzie.com. Not only that, they can also search and browse podcasts via Feedzie.com.

Last year, semoz.org gave Feedzie.com an honorable mention award in the context of Web 2.0.

As we serve to more and more people every day, we decided to improve the way Feedzie.com handles content and gave a magical spin to it. At this stage of its development life, Feedzie.com is more content centric than ever. In fact, we believe, it is one of the unique search engines on the Internet that provide more content-centric features than simple statistics-oriented features, such as cross-link counts.

Our Solid Goals

With its new feature set, we upgraded our *goals-to-accomplish* and came up with following 4 solid goals:

GOAL 1

Improve the way we understand topics mentioned in published content and present extracted information in such effective way that our users can easily make use of it at finding things that they are interested in.

Provide better and more service to all our visitors, especially to the users who are using our software applications, namely Ziepod and Newzie.

Advertise feed publishers by presenting their content to the right audience thru our unique content-centric features at Feedzie.com and provide tools for their websites.

Use Feedzie.com as a starting point for other new projects in near future by making use of gained knowledge and improved techniques.

To us, every project is a starting point for another project. Every another day, we will feed you and provide better service thru Feedzie.com. And when time comes, it will help us build another project on top of that. So, in a sense, Feedzie.com is a stepping stone for our future.

Last Words: Feedzie, not like any search engine you see

Finding things that you like in the Blogosphere is like searching needle in the Amazon(not with the .com). There are many search engines on the Internet that help you find things. Even though search results are mostly related, it doesn't mean that they show you the most related content. Nobody can guarantee that. At this point, we decided to look at this problem from different angle.

Unlike any search engine, other than search results, we are feeding you related content so that you can expand your search and find interesting things that you even don't think of when you start searching. After all, we want you to enjoy your time with the things that Blogosphere produces.

Have fun!