IKANGAI Solutions. e.u.

Mobile Business Solutions

T F G+ E

Posts Tagged ‘Public Paper Writing’

Motivating an Idea in a Scientific Paper

May 14th, 2012 by Martin No Comments

When writing a paper, authors often struggle with the application of their ideas in real life. It is often the case that an idea is interesting, but fails to convince others because of the lack of immediate applicability or a difficulty to provide real world data. This problem is faced by many authors; for example, in 1989 when Tim Berners Lee proposed an information system that later became the WWW, there was no proof that this could work.
So, in order to convince others, authors invent application scenarios that illustrate the application of their idea in a “realistic” setting. However, sometimes these settings are not well fitted and convincing. This was also one of the reasons for the rejection of our SOCA paper. So we thought of new scenarios and actually found them right away: we looked at our own marketing methods and we decided to describe them with with Tweetflows. The setting has completely changed and moved towards crowd sourcing without being a typical crowdsourcing example like tagging images. Instead, we are able to foster the ground for a new concept which we call social tuple space (STS) that is used to coordinate the services during the execution of the Tweetflow. Within this setting, we are able to describe a new working example as follows:

A startup company (ikangai solutions) is launching a new mobile app and as promotional offer they will give their customers a promotional discount, i.e., unlocking of additional App features for free, if customers invite the their friends to download the app. Customers can invite 5, 10, 20 or more friends from their social network to benefit from different promotions (bronze, silver, gold or platinum), with 5 being the lowest number for a promotion.
Customers can optionally customize their invitations before sending them to their friends. The invitation contains a link to a marketing survey which will be collected by the ikangai solutions to get feedback for the app (e.g., if the app was downloaded). Once the invited friends have completed the online survey, the customer receives the promotion code and can unlock new features in the App.
For the promotion, ikangai solutions uses its own Twitter network and posts a Tweetflow that describes the necessary activities needing to be executed:

LG didStart.Tweetflow #app_marketing
SR @cerridan download.App http://www.ikangai.com/ #app_marketing
SR @cerridan select.friends #twitter #app_marketing
[SR @cerridan select.Friend | SR @cerridan customize.Invitation | SR @cerridan send.Invitation | SR (Friend) confirm.Invitation | SR (Friend) complete.Survey http://wwww.ikangai.com | SR @cerridan send.Info] {5,} #app_marketing
SR @ikangai unlock.Bronze {5} #app_marketing
SR @ikangai unlock.Silver {10} #app_marketing
SR @ikangai unlock.Gold {20} #app_marketing
SR @ikangai unlock.Platinum {21,} #app_marketing

Conceptually, these commands create a social tuple space that contains Tweets of the followers of ikangai solutions. These Tweets represent information tuples that are used to coordinate the execution of the Tweetflow. The tuple space exists as temporary space: as long as the Tweetflow is executed, the Tweets are available. The Tweets are automatically purged from the social tuple space as Twitter only retains a history of about 3000 Tweets for each user. The creator of the tuple space monitors the execution by following the Tweets of the users:

LG @ikangai didSelect.Friend #app_marketing
LG @ikangai didCustomize.Invitation #app_marketing
LG @ikangai didSend.Invitation #app_marketing
LG @cerridan didConfirm.Invitation #app_marketing
LG @cerridan didComplete.Survey #app_marketing
LG @ikangai didSend.Info #app_marketing

LG @cerridan didUnlock.Bronze #app_marketing
LG @cerridan didFinish.Tweetflow #app_marketing

We believe that this example is illustrative and clearly motivates the use of Tweetflows in such settings. One of the next steps is to elaborate on social tuple spaces and to provide a simulation of Tweetflows in the Twitter Testbed of Andreas Scharf.

your ikangai science team

Translating Tweetflows into Natural Language

April 8th, 2012 by Martin No Comments

We’ve stared an experiment on Amazon Mechanical Turk that studies the translations of Tweetflow commands into natural language by “Turkers”. In the experiment, we briefly explain the Tweetflow syntax to the Turkers and ask them to translate several different Tweetflow commands into natural language:

Tweetflows are a simple syntax that are used to ask users on Twitter to perform certain activities.

  • Task 1: Translate SR @ikangai doProofread.Blogentry into an English sentence:

Example: SR @ikangai doWrite.Paper #Tweetflows #language-syntax tranlates into:

ikangai, can you write a Paper about the language-syntax of Tweetflows?

 

  • Task 2: Translate SP @cerridan proofread.Blogentry into an English sentence:

Example: SP @ikangai write.Paper #Tweetflows #language-syntax tranlates into:

ikangai, I can write a Paper about the language-syntax of Tweetflows.

 

  • Task 3: Translate the following four Tweetflow commands into four English sentences:

SR update.qcard http://www.ikangai.com//apps/qcard

SP @ikangai canUpdate.qcard http://www.ikangai.com//apps/qcard

SR @ikangai doUpdate.qcard http://www.ikangai.com//apps/qcard

SP @cerridan didUpdate.qcard http://www.ikangai.com//apps/qcard

 

  • Task 4: Translate the following sequence into two English sentences:

[SR @ikangai doPublish.Description http://www.ikangai.com/ #QRCode #Scanner #iPhone | SR @johannes2112 proofread.Description http://www.ikangai.com/ #QRCode #Scanner #iPhone]

 

Example: [SR @ikangai doUpdate.Description http://www.ikangai.com//apps/qcard-digital-business-card/ | SR @cerridan doAdd.Paragraph http://www.ikangai.com//apps/qcard-digital-business-card/]  translates into:

ikangai, can you update the description (see: http://www.ikangai.com//apps/qcard-digital-business-card/)? cerridan, can you add a paragraph after ikangai has completed (see: http://www.ikangai.com//apps/qcard-digital-business-card/)?

The goal of the experiment is to provide proof to support the claim that Tweetflows are more human readable than for example SOAP requests:

<?xml version=”1.0″?>
<soap:Envelope
xmlns:soap=”http://www.w3.org/2001/12/soap-envelope”
soap:encodingStyle=”http://www.w3.org/2001/12/soap-encoding”>

<soap:Body xmlns:m=”http://www.example.org/stock”>
<m:GetStockPrice>
<m:StockName>IBM</m:StockName>
</m:GetStockPrice>
</soap:Body>

</soap:Envelope>

your ikangai science team

PhD Thesis – Motivating Scenario

February 12th, 2012 by Martin No Comments

In our working scenario, we look at the creation of a presentation for a keynote at an archeological conference. The speaker, being a university professor, wants to give an overview on the current work of his group and highlight some of the recent findings. He decides to show the recent activities of his group with short videos from excavation sites during the keynote. Thus, the professor asks his co-workers to provide him some of their material so that he can include it in the presentation. Being an archeological group, several team members are working in the field and are not available in the office for a direct consultation. Thus, they need to be reached via email, text message or phone call. Remote team members also need to be reminded not to forget to produce their input for the presentation, because they do not necessarily have their laptops around when working on the excavation site.

After the collection of the video material is completed, the material needs to be edited according the needs of the professor for his keynote. For this purpose, the university provides a video editing service which takes the raw video material, a description of the desired outcome (duration, resolution, deadline) and produces a new video. This activities can be iterated over several steps to produce the final version of the video. Finally, the professor can make a reservation for a presentation room to practice his keynote to ensure that his laptop works with projectors and ask colleagues for feedback on his presentation.

If we treat the application scenario as Service composition/Service coordination problem, we can observe that there are several challenges that are difficult to address with existing approaches. First of all, we need to integrate different types of services into a Service composition. In our scenario, we have human provided Services like the recording the video on the excavation site and the video editing service. The location based reminder Service is embodied as App on mobile devices, as well is the video upload Service. The reservation consists of making a phone call and inviting colleagues to the trail keynote talk requires contacting them via email or phone. Secondly, the group working remotely requires asynchronous communication, because the not every person at the excavation sites may be reachable all the time. Thirdly, the inclusion of the social context needs to be addressed as well. Horowitz discussed this aspect as the village paradigm. In villages, knowledge dissemination is achieved socially: information is passed from one person to another person, and eventually the right person is found who is able to answer a particular question. In our scenario, we can make use of the very same paradigm, but our focus on the discovery of Apps. A typical conversation could be imagined as follows:

Person A: Do you know a tool to upload videos to our site?
Person B: Well, let me think \dots I know that C has a tool on his mobile that you could use. Ask him!
Person A: (asking C): B told me, you have a tool for video upload?
Person C: Yes, you can use my mobile to make the video and upload it. It also tweets the link to the video, if you like. But you can also download it from the App store. Just look for Video Tweeter.

As a consequence, we need to consider the social context during the discovery process. This requires us to address to additional challenges: a lightweight coordination language which can be used in the social context and support for monitoring in order to keep track of the progress of Service compositions for coordination purposes.

your ikangai phd writing team

(c) 2012 Martin Treiber – All rights reserved.

PhD Thesis – Introduction

January 23rd, 2012 by Martin 1 Comment

Principles of Service Oriented Architectures (SOA) have been successfully applied for the solutions of business problems and related challenges regarding the integration of heterogeneous software systems in business processes. For this purpose, the so-called Big Web Services technology stack with standards like SOAP, WSDL, WS-Addressing or WS-Security was devised and supporting tools were implemented. These standards ensure that there is a level of common understanding between businesses partners and allow for platform independent communication between remote software Services and their composition/coordination with dedicated languages like BPEL or YAWL.

However, due to the perceived complexity of Big Web Services, Restful Web Services that build on well known W3C/IETF standards such as HTTP, XML or URI have been proposed as (lightweight) alternative. In particular, so-called Service Mashups were introduced, which provide solutions to very specific and narrow business integration problems. The ability of non experts to create Mashup applications is considered to increase the productivity of employees in their daily routine work and when working with different companies on joined projects.

In this context, a growing number of Services can be observed that represent real-world business activities which are performed by humans rather than Software. A prominent example is Amazon Mechanical Turk where tasks are distributed to human workers (Turkers) that offer all kind of Services. As a result, additional standards for requesting work of humans via standardized Service interfaces have been devised, including WS-Human Task and Human Provided Services (HPS). These standards allow for the seamless integration of humans into existing workflows and foster the collaboration in so called mixed Service environments of human and software Services.

The parallel rapid development of mobile technologies and their wide spread adoption, led to a situation, where many employees actually do their job while they are moving. Thus, mobile employees often perform Service work remotely at the location of the customer and their work require the availability of mobile handheld devices such as tablets and smartphones. While early mobile Services focused on simple tasks of entering the data in a dedicated handheld device, modern mobile workers often face complex tasks that require the coordination of several activities of different mobile employees.

Consequently, a lightweight approach for Service compositions is also attractive for mobile workers, because existing SOA standards are difficult to implement on mobile devices because of the volatile characteristics of mobile devices in terms of network connectivity and availability due to limitations in terms of power supply. These are the main reasons that past efforts of adopting SOA on mobile devices focused on the consumption of remote (Web) Services from mobile devices, whereas the provision and composition of Services on mobile devices did not receive as much support. Instead, Apps have emerged as primary means for Service provision on mobile devices. Apps are independent pieces of software that offer a well defined set of functionality are controlled by the user. On average, users install approximately between 14 and 40 Apps on their devices that cover a broad range of functionality: social networking, weather, sports, location information, dictionaries, photography and games are among the most common classes of Apps being used.

In our work, we study the applicability of SOA principles with regard to existing (mobile) infrastructures like App Stores and mobile Apps. Specifically, we analyze established SOA principles like Service provision, Service binding, Service discovery and Service composition in the context of mobile devices and investigate the mapping of SOA infrastructure to a mobile infrastructure. We also address the social aspect of mobile Service provision: mobile Services – being bound to a human – run on devices which are controlled by the owner who can decide if a Service is executed (provided) or not.

To address core mobility aspects like limitations of connectivity, we introduce a lightweight programming language called Tweetflows that provides communication mechanisms to invoke Apps remotely and consequently the means to compose Apps. In this regard, we study the application of microblogging services (e.g., Twitter) as communication backbone for the use of mobile Services in a social context.

your ikangai phd writing team

(c) 2012 Martin Treiber – All rights reserved.

Writing a PhD Thesis

November 19th, 2011 by Martin No Comments

The process of writing a PhD thesis can be compared to finding a path through a labyrinth: if you are lucky, you find Ariadne’s thread and can follow it. If not, you are in trouble: you take a lot of (wrong) turns and come back to the place where you started and risk to be trapped for a longer while…
In my case, I believe that I have found Ariadne’s thread – after I took a couple of wrong turns. I’ll start as of today with writing my PhD thesis on the blog. More specifically, I’ll set up a wiki in the next days and will add material there over the next couple of months.

So, you might ask, what is your PhD thesis all about? I’ve actually managed to say this in one, simple sentence (with a lot of technical terms):

I investigate the application of SOA (Service Oriented Architecture) principles on Social Networks for the execution of lightweight Service based workflows.

SOA principles are methods that are used to create Service Oriented Architectures and describe the way such systems are designed and built. For example, Services are implemented in such a manner that they can be discovered, bound and then executed by the Service requester. Well, THIS is certainly a no-brainer, but you wouldn’t believe how big this problem for the Service community is and how many papers on this have been written in the past :-) . I won’t go into details here, but yet alone the notion of Service is not well defined and remains subject to discussions. For my work, I refer to the work of Daniel Schall who did a good job explaining Services and in particular Human Provided Services which I use in my work.

The term “lightweight” is used for describing something that is not the “real deal”, but rather some very specific down scaled thingy :-) . I do the same with workflow languages: I do not bother to implement another complex workflow language, but focus on some basic constructs that allow for the creation of simple descriptions of work that can executed in a social network. Since people are not good when they need to do complex stuff, I think I am on the safe side here :-) . Of course, I’ll elaborate on this a bit more in my thesis :-) .

For now, my thesis consists of four parts, which organize my contribution into four areas, which build on each other and will be explained in the next blog entries:

  • social network
  • service
  • communication and coordination
  • monitoring

[UPDATE - The wiki that will contain my PhD thesis is online]

your ikangai phd thesis writing one man science team (yiptwomst)

Creating contextualized mobile workflows

August 8th, 2011 by Martin No Comments

This post continues our work (introduction , application scenario) on mobile workflows which we conduct during a public paper writing process.
Context information is considered as important additional information source for workflows during their execution. Existing approaches use context information (location, time, user profiles) during the execution for various purposes, such as adaptation. All approaches need to specify the type of contextual information during the design time which are bound during the execution of the workflow. However, in situations, where the format of context information cannot anticipated a-priori, we need to follow a different approach. This is illustrated by our working example, where context information can be available in different formats: pictures, geo coordinates, textual descriptions or videos. Furthermore, the availability of this contextual information may even be discovered directly at the site. For example, a picture of an entrance gate with a person standing next to it can be used to assess the size of the gate. Thus, it is very difficult to It’s very difficult to create generic context slots because of the arbitrary nature of context information. To overcome this limitation, we propose to integrate concrete context data the design of the workflow. In doing so, we use context information where it is available: at the location where the workflow is designed.

Example. The power generator delivery service must know where to deliver the power generator and requires information about the location infrastructure. Thus, a request a posted, to collect pictures of the festival location and to upload them to a picture sharing site to make it available for the delivery company. A video of the organizer is made explaining the for security that the presenter of this video is allowed backstage access to install the power generator. A code snippet of this workflow is shown below.

LG didCreate.TwitterList http://festival.org #powergeneratordelivery
SR @security follow.TwitterList http://festival.org #powergeneratordelivery
SR make.Pictures location=festival | SR upload.Pictures url=http://festival.org #powergeneratordelivery
SR comment.Pictures url=http://festival.org #powergeneratordelivery
SR @delivery.folllowTwitterList http://festival.org #powergeneratordelivery
SR @delivery download.Pictures http://festival.org
SR @delivery notify.arrival @security
LG didUpload.Video http://festival.org | SR @secretary email.Video @delivery

In a first step, we create a Twitter follower list that severs as black board for the exchange of messages concerning the Tweetflow. This makes it easy for all participants to follow the Tweets that are exchanged in the context of the delivery of the power generator workflow. Context information (pictures, video) is directly linked in the Tweets of the Tweetflow. It is worth noticing, that the actual interpretation of the context data depends on the receiver of the message. In our example, context information such as pictures or video are interpreted by people and people provide the actual service.

your ikangai science team