You can now filter results from the uberVU conversational engine based on language. We noticed getting results from all over the globe can become quite chaotic so language detection was a must.

uberVU is now smart enough to understand:

• arabic
• bulgarian
• chinese
• danish
• dutch
• english
• persian
• finnish
• french
• german
• hungarian
• icelandic
• italian
• japanese
• korean
• norwegian
• polish
• portuguese
• romanian
• russian
• serbian
• spanish
• swedish
• thai
• turkish
• ukrainian

This brings a new level of refinement to the Context Voice API and you can now use it for localized search results. We are looking fwd to see what you can build.

Read the full documentation.

Today is party time at ContextVoice / uberVU as we’re officialy launching our Search API as part of the ContextVoice suite of APIs.

Why another Search API? Don’t we have blog search and Twitter search already?

Yes, we do. But we do something completely different. We do conversational search. A conversation is usually a story (blog post, news article) and all its related comments from all over the Web. So a blog post that has comments on Digg, has been retweeted by people and has got comments on FriendFeed is, for us, a single conversation.

Blog search only indexes blog posts and leaves out comments of any kind. And it’s usually slow.

Twitter indexes only tweets that contain the keyword you’re searching for. It’s close to real-time.

We index both the story content and all the distributed comment content and we return as a result complete conversations. And we’re close to real-time. The best of both worlds.

And this is useful because…?

The Web right now is all about conversations. However, until now, search engines have mostly returned results by relevance (Google), recency (Twitter) or some sort of combination (blog search). But out of those recent results, which one is more important? Where’s all the attention? What’s hot right now as opposed to what’s just fresh?

What’s very important to know is what Conversations (not blog posts) are HOT right now – so you can participate and make decisions in close to real-time. That’s what our API is focusing on.

What to expect from our search

Full conversation search. Because we don’t index just page content but all the reactions and comments from all over the web our results offer a better understanding of the conversation than a regular blog search. The blog post may be about the Google OS, but the comments are mostly about Microsoft. Our Search makes that visible to you.

Close To Real-Time results. We monitor the top social media sites in realtime so if something happens there, you know instantly.

Filtering options. We can show you the most relevant, the newest or the hottest conversations. We calculate the hotness based on conversation acceleration – how many reactions it got in the last time-frame. So a conversation with 10 reactions in the last minute may be hotter than one with 1500 reactions distributed over a month. But the hot conversation matters more, because that’s the one that has the attention and momentum.

What to build with the API

The Search we provide can be used in a lot of products and dashboards. uberVU.com, for example, is entirely built on top of this API. You should check it out as it’s a great starting point for what’s possible with the Search API.

Social media dashboards – want to search entire conversations for your brand of product and get an insight into what’s hot right now and where the attention is – whether on Twitter, FriendFeed or blogs? The Search API is a great place to start

Memetrackers – interested in building a meme site for the NFL or the latest gadgets? Just use the Search API and get the hottest conversations about that topic in close to real-time.

Community management projects – want to bring fresh content and community conversations from all over the Web to your site or community? Search for the topics you’re interested in with our Search API and expose the hottest conversations on your site. We don’t just get the stories, we get the comments. And the comments are the most important part of a community. People comment if other people have commented. That’s what you want.

Financial analytics – search for your favorite company and graph comments to the evolution of the stock quote. You may be intrigued at what you find.

These are just some starting ideas, I’m sure there’s a lot more you can think of.

We are moving some data around so the system will be down for 3 hours. We will be back around 2 PM GMT. Sorry for the inconvenience.

Later edit: Done

We had the pleasure to meet the Traveling Geeks in London today. It was super-cool to meet people like Robert Scoble,  Craig Newmark, Howard Rheingold, Meghan Asha, JD Lasica, Susan Bratton, Jeff Saperstein, Renee Blodgett, Jim “Sky” Schuyler, Ayelet Noff and to get their thoughts on what we are building. Awesome Seecamp event.

It was a very unusual networking event because we had a 3-minute pitch for the geeks to get an idea on what we are doing and then talk to each of them for 9 minutes. If you are thinking of speed dating you are right. That was exactly how it was. In our pitch we decided to talk on why building the social graph is important. Main takeaways:

• We are talking about conversations for more than 10 years now but all that we have is millions of unrelated messages, with no context that you can not do anything interesting with them.
• Context Voice brings order to all the chaos by creating a search-able index of platform agnostic conversations that you can track in real time.
• This is very important because we help people find and interact with their tribe.
• You can do all sort of amazing stuff on top of all the data we are giving you.

Robert also did a nice live interview with us that we embedded bellow:

Related articles by Zemanta

• Traveling Geeks are Coming to London (blogs.msdn.com)
• Traveling Geeks Are Going to London! (blonde2dot0.com)
• Traveling Geeks Are Going to London! (thenextweb.com)
• Coming Soon: Even More TechCrunch in London (techcrunch.com)
• Introducing Traveling Geeks London (socialmedia.biz)

We made some updates to our pricing/usage scheme today:

• Commercial use for the free version
  Our thinking was that if people make money out of the API it would be nice to get a share of that. What we didn’t take into consideration was that you can still start a commercial project and not make any money for some time.  Not to mention that many of the applications that people thought of building with the API were commercial apps. So from now on you can use the free version for all your business projects too.
• Number of API calls
  We still have the 5000 calls/day limit. However if you are working on a cool project and need more than 5000 calls give us an email/ reach us on Twitter etc, tell us about it and we would be more than happy to increase that limit to help your app grow.

Have fun and enjoy the summer!

Related articles by Zemanta

• The Social Data Player Platform is Now Commercially Available – Embed Datasets in Your Website (socrata.com)
• Twitter Increases API Limit (mashable.com)

Version 1.1 of the Context Voice is out. And we have many things to announce:

New calls and parameters

You probably noticed it takes a long time to get a conversation that has many thousand reactions. We did a few things to improve this issue:

1. We know have 2 endpoints: one for reactions (where you just get the comments for an URL) and one for a resource (that you can use to get meta-data about a URL-based conversation). So for things like title, content or generator of a URL you know have a new endpoint called at http://api.contextvoice.com/1.1/resources
2. We introduce paginated responses – if you read the documentation you will see two parameters: page and perpage that will help you achieve this. You no longer have to download megabytes for one response
3. We automatically track top-blogs. And now we update them faster.
4. You can now order the reactions asc/desc using the ‘created_at‘ field

New platforms

We now introduce 4 new platforms where we gather comments/reactions from: StumbleUpon, Delicious, Mixx and NewYorkTimes. What this means is than from now on we also get comments from NewYorkTimes website and for all the links in our index we get comments from StumbleUpon, Delicious, Mixx.

Simple Update Protocol

We now support Simple Update Protocol or simply SUP. SUP is a project developed and maintained by the good folks at FriendFeed. Context Voice SUP is a a simple and compact “ping feed” that alerts your web app when a conversation has been updated. This reduces update latency and improves efficiency by eliminating the need for frequent polling. Not so many calls means less bandwidth and for paying customers money savings.

Context Voice SUP is available at http://api.contextvoice.com/sup.json/?apikey=yourapikey

Non-authenticated calls

You can now make non-authenticated calls to get reactions for a given URL. So if you only need to get comments for a given URL use http://externalapi.contextvoice.com/1.1/reactions/?url=http://www.yoururl.com

You are restricted to 500 anonymous call per day per IP. This way of using Context Voice API is perfect for plugins, widgets or any type of app that’s distributed on users machines rather than running from one server with one IP address.

That’s about it… Let us know what you think. We welcome any feedback.

Anything else? We also have a new page with wrappers: a Ruby one and a PHP one. Many thanks to the wonderful people who developed them.

PS 1: “What about search?” you may ask. We know… we know. It’s coming but we need a little bit more time to polish and document it. Bare with us and you will not be dissapointed.

PS 2: version 1.0 is still supported by the way. You can even use the new platforms with it. But we recommend you switch to the new version asap.

Related articles by Zemanta

• StumbleUpon’s Su.pr URL Shortening Service Is Now In Private Beta (250 Invites) (techcrunch.com)
• Mixx Still Growing (But Not Fast Enough), Relaunches Website (techcrunch.com)
• 5 Essential Steps to Make Your Site Search & Social Media Friendly (mashable.com)

We are running some database maintenance jobs between 4 AM and 6 AM EST. The service will still be available but some resources will be unavailable for 2 h. Thanks for your patience.

Later edit: Finished. Back on track!

Today, at Open Coffee Bucharest, we presented Context Voice publicly in front of a 25-people crowd. The feedback we got was awesome and we thank to all the people who came to ask questions and see how we can help them.

Our presentation is below:

The main ideas/takeaways from our talk:

• We add another layer to the social stack. After the conversational graph (Facebook) or message graph (Twitter) we want to offer a conversational graph.
• It’s a service that you can use through an easy to use API
• Main features of the API are search, filtering, SUP support and close to real time URL tracking
• You should use our API because you save time and money
• You can do tons of stuff on top of it: from blogging tools to financial applications or meme websites.

Few pictures from the event:

Context Voice Logo

We are happy to announce that, as of today, the uberVU API will be called ContextVoice and will live at ContextVoice.com.

For now, apart from the name change, there are no other changes that have been introduced. So you might ask then, why do it?

The reasons are many.

First of all, the uberVU API seems to have taken on a life of its own. Lots of people want to use it for things we never would have imagined. When we tell people what we do, the first question the ask is usually “Do you have an API?”.

Because people want to use it to do very different things than we have originally thought, we’ll have to add things in the API that have no relation the uberVU as a product. What uberVU does right now will, in the future, only be a small part of what ContextVoice will offer. That’s one of the reasons why we felt the API needed to be a product on its own.

Secondly, while uberVU is still in the Private Beta works for the time being, the API is now public and people are willing to pay for it and use it commercially. This subtle difference between where the product and the API is in terms of development and reliability was not apparent from the uberVU site. So it confused a lot of people that wanted to find out about the API and use it within their apps.

For now, you’ll see no change in the API calls made if you’re already using the uberVU API. If you’re thinking of starting to use it, however, we suggest going to the website and start using ContextVoice.

We’ll announce changes made to the API as they happen, making sure it’s not a hassle for you to implement those changes, should that be the case.

As, always, we’re very interested in knowing what you think about this. Shoot your thoughts in the comments, on FriendFeed or Twitter.

We have a few updates to the uberVU API that we think we’ll help you create even better mashups.

More platforms
Today we introduce 2 new supported platforms: Reddit and HackerNews, two very popular websites for Internet savvy people (aka geeks). You don’t have to do anything special to support the new platforms. We just added two more items to the generators list.

Faster reaction gathering
Twitter, Digg, Reddit and HackerNews are close to real-time. In matter of seconds/ minutes every reaction/ comment published on one of those networks will become part of a conversation that you can follow through the API.

Support for (almost) all URL shortening services
We now get all the links sent on Twitter. As long as the URL shortener through which the link was sent supports redirect, we can get those initial URLs for you.

Ruby wrapper
If you use Ruby, Radu Spineanu was kind enough to code a Ruby wrapper that you can use to jump-start your mashup.