Thursday, September 26, 2013

Home automation standards

Wednesday, September 25, 2013

How do you explain Machine Learning?

It looks like the marketing won in math too. As per this thread How do you explain Machine Learning and Data Mining to non Computer Science people? any statistics is machine learning. As I can guess, everything unknown is a deep learning ...

Tuesday, September 24, 2013

Search for online courses

Our mashup based on Google Custom Search Engine - Search for Online Courses. And discuss your result right on the page with TogetherJS from Mozilla.

Monday, September 23, 2013

Sunday, September 22, 2013

How Twitter can cash in?

MIT review published an article about the possible earnings for Twitter.

More or less traditionally referred to in conjunction with Twitter sources. At the first hand it is location based services. Although fewer than 1 percent of tweets are "geotagged," or voluntarily labeled by users with location coordinates. The second big area is natural language processing for tweets. For example, there are various demographics data that could be extracted and making sense of breaking news.

We would like to point attention to our old idea - Twitter as a transport. The system could be used as a transport layer. Think for example about many services that provide request/response cycles over SMS. Why do not use Twitter for this? Request data (and/or service) over Twitter rather than SMS.

For example, try to send the following tweet:

@t411 t GOOG

It will return to you (as a reply in Twitter) stock quotes for GOOG. Here:

@t411 is service address (a-la service number in SMS world)
t - is a request
GOOG is a parameter

It is from our T411 for Twitter service. And it is DIY service. You can define your own bot at T411.

Obviously, that such service could be a part of Twitter's offering. E.g. allow unlimited statuses for a fee, etc. It is Twitter for business.

Wednesday, September 18, 2013

How to replace Apple's iBeacons

Explanatory article describes iBeacon from Apple.

iBeacons is a Bluetooth-based micro-locations system. But instead of being used by people to determine their own locations, it's used by retailers, museums and businesses of all kinds to find out exactly where people are, so they can automatically serve up highly relevant interactions to customers' phones.

How does it work? The closes analogue is, probably, an automatic check-in. As per Apple, if you walked into, say, Jay's Donut Shop, iBeacons would know for certain that you had walked into Jay's Donut shop, whereas other location apps might use GPS, Wi-Fi and cellular triangulation to produce a list of guesses about where you were. A check-in wouldn't even be required.

But of course, it depends on pre-installed BLE devices (iBeacons). They have to have some global addresses (unique IDs) in order to distinguish Jay's Donut Shop from Ann's Donut Shop.

And here I would like to highlight again our old idea about triggering data access depends on the network environment. It is SpotEx. See our papers, for example. In this concept, any existing or even specially created wireless network node could be used as a presence sensor that can open (discover) access to some dynamic or user-generated content. The content itself could also be linked to social media. An appropriate mobile service (context-aware browser) can present that information to mobile subscribers. Potential use-cases for the proposed approach include any project associated with hyper-local news data. For example, projects providing Smart City data, delivering indoor retail information, etc. In other words, we can replace iBeacons right now (more precisely - simulate the same behavior) with Wi-Fi nodes. And because Wi-Fi access point could be opened right on the mobile phone, any smartphone can play a role of iBeacon.

Actually, we wrote about this in discussion about Estimote. Once again - any wireless node (e.g. Wi-Fi access point or even the smartphone itself) is a beacon. The location is completely insignificant here. It is about the visibility only. As soon as some access point is visible (and this access point could be opened right on the phone, of course), we can deliver some data to the mobile user (to the subscriber).
Of course, the metric could be more complex (e.g., we can use The Spearman rank-order, etc.), but the whole idea is transparent. The presence statement for some network node (nodes) triggers data access.

Tuesday, September 17, 2013

M2M & IoT

In short, 'M2M/IoT Application Platforms' represent M2M platforms re-cast for the age of the Internet of Things.

The M2M/IoT Application Platform provides the 'glue' that intermediates between application developers, M2M connected devices and a range of niche and specialised M2M platforms and wider enterprise IT systems. Referring to the dynamics of this new M2M/IoT world, Morrish commented: "In the world of the M2M/IoT Application Platform, the application developer is king." - from here

Monday, September 16, 2013

The Machine Learning Summer School

Slides and tutorials from The Machine Learning Summer School
26 August to 6 September 2013 at the Max Planck Institute for Intelligent Systems, Tübingen, Germany

Saturday, September 14, 2013

Emergency Indoor Navigation

How New Indoor Navigation Systems Will Protect Emergency Responders. Tracking firefighters in blazing buildings helps keep them safe - an interesting article.

Thursday, September 12, 2013

Twitter Data Analytics

Free book: Twitter Data Analytics.

This book is designed to provide researchers, practitioners, project managers, and graduate students new to the field with an entry point to jump start their endeavors. It also serves as a convenient reference for readers seasoned in Twitter data analysis.

/via Data Central

Wednesday, September 11, 2013

INJOIT: correction

Thanks to Justin Hill from CrossTalk Publisher. He pointed us to an invalid article, which we rashly published in INJOIT. It is our fault and it has been fixed. You can see an updated issue 6 here.

Tuesday, September 10, 2013

Location Sharing Without Central Server

Local Messaging

Our missed paper:

Dmitry Namiot, Manfred Sneps-Sneppe. "Local messages for smartphones". Future Internet Communications (CFIC), 2013 Conference on, pp. 1-6, DOI: 10.1109/CFIC.2013.6566322.

This paper describes a new model for local messaging based on the network proximity. We present a novelty mobile mashup which combines Wi-Fi proximity measurements with Cloud Messaging. Our mobile mashup combines passive monitoring for smart phones and cloud based messaging for mobile operational systems. Passive monitoring can determine the location of mobile subscribers (mobile phones, actually) without the active participation of mobile users. This paper describes how to combine the passive monitoring and notifications.

Monday, September 09, 2013

INJOIT: call for papers

The International Journal of Open Information Technologies (INJOIT) is an all-electronic journal with the aim to bring the most recent and unpublished research and development results in the area of information technologies to the scientific and technical societies. Free, peer reviewed papers.

It is published by the OIT Lab (Open Information Technologies Lab, Faculty of Computational Mathematics and Cybernetics, Lomonosov Moscow State University).

Wednesday, September 04, 2013

Tuesday, September 03, 2013

Campus Party Europe

We continue to share links for monitoring the interesting events in Twitter. Now it is Campus Party Europe, London 2013

Campus Party Europe

/via Geo Messages

Monday, September 02, 2013

Trajectories and proximity

A new paper: Dmitry Namiot "Flock Patterns and Context", Applied Mathematical Sciences, Vol. 7, 2013, no. 90, pp. 4493 - 4497, HIKARI Ltd

The wide deployment of location detection devices (for example, smartphones) leads to collecting of large datasets in the form of trajectories. There are a whole set of papers devoted to trajectory-based queries. Mostly, they are concentrated on similarity queries. In the same time, there is a constantly growing interest in getting various forms for aggregating behavior of trajectories as groups. The typical task, for example, is find all groups of moving objects that move together. For example, we can find convoys of vehicles, groups of people, etc. In this paper we discuss the task of flocks discover y for context-aware applications, where location data could be replaced by proximity information. We propose a framework and several strategies to discover such patterns in streaming context-related data. Our experiments with real datasets show that the proposed algorithms are scalable and efficient.