Recommendation System for Information Services Adapted, Over Terrestrial Digital Television

Computer Science & Engineering: An International Journal (CSEIJ), Vol.2, No.6, December 2012
DOI : 10.5121/cseij.2012.2602 17
RECOMMENDATION SYSTEM FOR INFORMATION
SERVICES ADAPTED, OVER TERRESTRIAL DIGITAL
TELEVISION
Mery Y. Uribe Rios1
and Rafael V. Páez2
1
mery.uribe@javeriana.edu.co
2
paez-r@javeriana.edu.co
1,2
Pontificia Universidad Javeriana, Bogotá, Colombia
ABSTRACT
The development of digital television in Colombia has grown in last year’s, specially the digital terrestrial
television (DTT), which is an essential part to the projects of National Minister of ICT, thanks to the big
distribution and use of the television network and Internet in the country. This article explains how joining
different technologies like social networks, information adaptation and DTT, to get an application that
offers information services to users, based on their data, preferences, inclinations, use and interaction with
others users and groups inside the network.
KEYWORDS
Application, social network, services adaptation, terrestrial digital television, Ubiquitous computing, web
services
1. INTRODUCTION
The development of apps for digital television in Colombia is a starting process, due to the state
of this technology in the country and to the projects of Ministry of ICT to the expansion of DTT,
with the aim of replace the analogical television, which currently is over the most part of the
national territory [1] [2]. Because of this, the implementation of applications of this type
represents an opportunity of innovation, participation and research in this context.
Being a part of this innovation, we must participate with original ideas that would bring benefits
for the finals users and for the designers, and developers of those apps. We are trying to enclose
these issues with the outgoing interaction user data in the network in DTT, in order to adapt
services according to the specific needs of each user. All this, to take advantage of the social
network boom, the benefits of services adapted and the big distributed network deployed for all
users of DTT.
This document explains the design, development and implementation of a social network for DTT
used as a recommendation system, based on the interaction among its users.

18
2. SOCIAL NETWORK FOR INFORMATION SERVICES OVER DTT
2.1. Principal Base Architecture
The interactive app for DTT is part of a project called “Architecture of Adaptation of Information
Services over Digital Television, supported on a social network”. With the implementation of this
app, we want to validate some aspects of this architectural design and make of this a study case.
The architectural styles to validate through the application are: separation of the composition,
design and logic of an app, client – server, architecture for layers, the use of a demand service,
separation of files for server and separation of services.
The validation was made through the implementation of the app over an environment of digital
terrestrial television, with a publicity recommendation system based on the interaction of
television users in the social network.
2.2. Adaptation of information
Based on system requirements and some services, which were taken as a study case to materialize
and exemplify a possible service for offering in the architecture, such as i) To offer a publicity
recommendation and ii) To notify a recommendation’s state, we defined a model of adaptation
based on profiles for enhance services due to their specification and to the needs of users, through
the management of some determinate data. The use of profiles lets to the user or group data
process , locating and using specific information for each case, as is mentioned in [3] “all the get
data must be integrated to create complete profiles of the consumers”. In this way, we can
represent preferences in a orderly, automatic way [4–7] and we can to find relevant information
which can to be used through processes like user profile mining [8] or recommendations services
[9]. It should be noted that this adaptation model is not limited to the services mentioned, they are
just a study framework; the model is general and transversal to any service.
We worked with five types of profiles, according to the kind of information to manage in each
one. The profiles were:
1. User profile: Stores information about basic data and user preferences.
2. Context profile: Uses information about events that involve the interaction of user and the
system [10]. Those data were needed to monitor some variables that can influence, change
and/or improve the services offered by the system. Also, these data were taken from the
television’s information system and the monitoring the system itself.
3. Device profile: Specifies the information related to the access device. These data were taken
in count in order to adapt the information and services, according to the respective features of
each device, particularly in some ones with limited capabilities, as mobile devices. These data
will be taken from CC/PP file CC/PP [11] of each device.
4. Social profile: Specifies the interactions of a user into the social network. These data describe
how a user is related to others users, group(s), services provided by the system and with
information about television content. All these behaviors are the most common patterns of
interaction recognized. Those data were taken from the interactions of each user into the
social network.
5. Group profile: Describes the data of a group of users, it can be generated by the system or
created by the user. We want to store management data about the group and its different
relationships with services and users. In this way is possible to work with each preferences’
group in the same way that the user’s ones. In this way, the services can be guided to a set of
users with some features according to specific preferences of each group. These data will
depend on the way them were created in the app (for the system or a user).

19
The adaptation model will be given for the follow figure:
ADAPTATION
User profile Device profile
Context profile
Group profile Social profile
Figure 1. Adaptation model
This figure shows the dependence of group and social profiles with the user’s one, and context
profiles to relate to the users. The creation and generation of groups and the interactions among
users will depend on the correlation between data of the profiles to find the necessary information
to offer services. It should be noted that the services to adapt not only depend on the social and
group profiles, but also to link information from all profiles.
2.3. Social Network
Some aspects like interactions among users, nodes network distribution, its representation and
how to analyze its data are taken in account for the network design.
According to the options of network data management explained in [12], is recommendable to
represent the network data like a graph, because it allows handling them in a generally and
characteristically than a sociogram (representation of a social environment, where the nodes or
points represent people, their goals or actions, while the connections represent the sequence that
connects them [12]).
Each interaction of a user (node) into the network will be represented like an arc in the graph,
depending on its type and its correlation with other data. These nodes are characterized by two
aspects: user and access device. In the last time, and depending on the reason of the relationship
among nodes, there are two kinds of interactions:
- Interaction of use: Related to the data profiles of each user and his activities into the network.
These interactions represent relationship between user – user, user – group and user – service.
- Interaction of resource: Related to technical aspects of the access device. These data will be
used to create interactions of resources, according to the technical capabilities and resources that
each node needs to deploy a service.
In order to identify groups and opportunities to offer services, central points in the network are
defined, which are points with the highest degree (highest number of arcs related (interactions)).
This analysis allows to create an interaction among nodes, for instance, an interaction among
users or among nodes for resources.
For finish this part, the distribution services method was defined, which will be “snow ball”
technique. This technique allows using the relationships among users like a “clinch” to nominate
to other users to a service or to nominate a service to a user [12]

20
2.4. Technical Aspects
To the implementation of an interactive app for digital television, we have taken into account
some technical issues, especially those related to the digital television standards DVB-T y DVB-
MHP [13–15], because they frame the activity in this context.
In this section the principal issues about the creation (design and implementation) used for the
app is presented.
2.4.1. User interface
One of the relevant aspects in the design and development of an interactive app for digital
television is the user interface, because the access devices and the execution environment are
different compared with an app for computer or mobile device’s one. The issues more outstanding
are those related to the navigability and the design patterns.
2.4.1.1. Navigability
This aspect frames the principal tips of the look and order of an app. The app control is specified
through the use of navigation’s buttons of a remote control, to manage the form and content of a
menu.
In first place, the remote control (see Figure 2) is the device by which the user interacts with the
television set and therefore, with the apps deployed in it. These remote controls have, generally,
four colors buttons (red, green, yellow and blue), that together with navigation arrows, are used as
controls for an application.
Figure 2. Remote control of TV set
Generally, each color key is associated with a determinate action, owing to the relationships of
the user with the different colors [16], the
Figure 3 shows the principal actions for each color.

21
Figure 3. Actions for colors [16]
In second place, is important to be carefully managing the menu content and navigation; the
information must be clear and concise, there should be no excessive submenus and should be easy
to move back and forth through them.
So, we reach an important point for the user interface of this type of app, the access device. As
this device can be a TV set, a mobile device or a computer, we can’t extend the content or the
navigation map. Although, we use the adaptation for taking in account some technical features to
the access device, we have to use a little more for content adaptation, so the app must be designed
to be deployed in any device.
2.4.1.2. Designed patterns
According to the application kind and the information amount to handle, the principal aspects can
have different behaviors. These aspects are:
• Application location [16] (see Table 1)
Table 1. Location aspects
Location Amount of
information
displayed
Size Location in the screen
Superposition Little information
related to content
showed in the
moment.
As small as
possible.
Bottom or top of the screen.
No cover logos.
Full screen
with video
So much content Use ¼ of size
screen for video
in the most
applications.
The video must be located
on left top of the screen.
Full screen
without video
So much content.
Use of complex
functions not related to
the TV program.
All screen All screen

22
• Menu [16]: to provide access to content and functionality (see Table 2).
Table 2. Menu aspects
Location Information
deployed
Use of content Highlight of content
Simple menu Use 3 to 8 menu
options
Select short and
accurate titles
for each menu
Provide a preview of the
option, for instance a text
or an image.
Mark the selected
element.
Multi screen Use for app that
provides multiple
video sources.
Use 3 to 8 videos
sources.
Play the sound
of the video
source selected
Mark the selected
element.
Index Select a structure like
an index that allows
to access in a quickly
way to popular
elements
The application
contents, some
contents and
options
Mark the selected
element.
Tabs Use them just for
users who work on
Internet.
Use horizontal
tags.
Select short and accurate
words for the title of each
tab.
2.4.2. Design and implementation
Having the mentioned aspects and specs on the system architecture, the user interface of the app
was designed and deployed. Some tools to do this process were used. Icareus iTV Suite 3.9 [17]
was used for the graphic interface, which “is an intuitive and versatile visual content development
tool which enables rapid development of digital TV content” [17]. This app allows to the user
interface to integrate external components, which are building by a software programmer, to give
new functionalities to an app , according to the needs. The principal activities were:
2.4.2.1. Design
The main displays (see Figure 4) of app was designed according to the topics about designed
patterns and having in account the important information to show in each one, because, it is an
application which is deployment when a TV content is transmitted.
As this application is highly interactive, different aids were used, which helped to this interaction
to be easier for using. For instance, some dialogs or a virtual keyboard were deployed, because
the remote control’s numeric keyboard of the TV set is not comfortable for the text exchange (see
Figure 5 and Figure 6).

23
Figure 4. User interface
Throughout the application we can see the designed patterns mentioned, for instance, the use of
the actions of each buttons color (see Figure 7), or the preview of the menu and sub menu options
(see Figure 8).
Figure 5. User interface - Virtual keyboard
Figure 6. User interface – Dialog
__________
Figure 7. Examples of control bars

24
Figure 8. User interface - preview of each option
2.4.2.2. External components
The external components are developed in an individual and personalized way according to the
needs of each application. The use of these components offers big benefits to the user through the
application handling, improve it with specific actions that make it adaptable to the required
behavior. These external components are development in a programming language and later they
are integrated to the user interface.
The main external components, according to the system requirements of the study case (publicity
recommendations), are:
• Principal: This has to give the access and load the user data (profile and photo).
• See friends: This has to load the data (profile and photo) of the current user’s friends.
• See groups: This has to load the data (profile and photo) of the current user’s groups.
• Search a person: This has to do the search of a person by his name.
• Search group: This has to do the search of a group by his name or topic.
• Do recommendation: This has to list the user’s publicity recommendations, confirm the
user whom is sending the recommendation (which have to be friend of the current user)
and send the recommendation.
• Recommendation: This has to load the recommendation information (Title, content).
With this principal functions list, the next step was their development. To do that, it was
necessary to count with others tools dedicated to the development.
1. For the development of external components: Eclipse IDE (development in java language)1
with Osmosys Plugin.
2. For web services development: because of the styles of architecture, the separation of the
composition, the design and logic of an app, because it is an app client-server and the
service separation, it was necessary to do the processing through of web services. In his
way, that app or the access device (which usually doesn’t have good processing
capabilities) don’t do the hard part of processing of the system, so its job is request and
receive the needed information. For develop them, it was necessary to have Eclipse IDE
(development in java language) with Axis22
Plugin and Apache Axis23
server.
1
http://www.eclipse.org/
2
http://axis.apache.org/axis2/java/core/tools/index.html
3
http://axis.apache.org/axis2/java/core/docs/installationguide.html

25
3. For the data base development: to design and deploy the used data base, we used MySQL
Workbench 5.24
and a MySQL 5.5.165
server.
Working with these tools, it is possible to develop the external components and request a web
service which uses the data base. Then, each component in Icareus iTV Suite 3.9 was configured
in order to use them inside the application.
2.5. Integration
Each external component is responsible for correlating profile data and those recollected from the
interaction among users into the social network. The component related with the
recommendations, help to reinforce this correlation with the association rules defined like method
to do the recommendations (see Table 3).
Table 3. Association rules
4
http://dev.mysql.com/downloads/workbench/
5
http://dev.mysql.com/downloads/mysql/

26
In these association rules, the recommendation type can be: 1: Academic-elementary, 2:
Academic-high school, 3: Academic-undergraduate, 4: Academic-Postgraduate, 5: Sport, 6:
Esthetics, 7: Movies, 8: News, 9: Information, 10: Food, 11: Music, 12: Religion, 13: Adults, 14:
Health, 15: Home, 16: Technology, 17: Child games, 18: Teenage games, 19: Culture, 20: Events,
21: General academic, 22: Babies clothing, 23: Boys clothing, 24: Girls clothing, 25: Teenagers
clothing, 26: Men clothing, and 27: Women clothing.
Having the external components deployed into user interface, the next step was to set the
complete app on the television server. To do this, we integrated the app to the application server
using Icareus iTV Playout, tool which manages the content of the TV server through channels and
its content. For this, it is necessary to have the app finished, the each external components
configuration files and set them into the server and deploy them on a display channel. After this
process, the application can be deployed and accessed from a decoder connected to apps server.
2.6. Validation
As the developed application is a part of practice validation of the architectural designed
mentioned, its implementation has the goal of verify the adaptation model, specially
implementing user profile, group profile and social profile, the social net model and the
architectural styles: i) Client- server, ii) Architecture by layers, iii) Separation of the composition,
design and logic of an application, iv) Sessions management, v) Services separation, vi)
Management, publication and distribution of information, vii) Management of resources
information and viii) Light client. The results are presented below.
2.6.1. Results
The final client is a small app with the minimal resources needed to execute it, complying with
the “light client” required.
The final app is made according to the layers of the architecture and its performance is correct in
respect of the main models.
The adaptation and social network models were applied with data of the mentioned three profiles
and the interaction layer was made under the association rules developed according to the
relationships between the designed profiles. These recommendations were applied and
transmitted from user to user, according to the defined rules of social network. This shows that
the aspects to validate with this study case were well in implementation and in deployment, for
this reason, we hope a good performance of the app with all aspects of the architecture.
Between the aspects to have in account in the implementation of this type of apps, are the tools to
develop the external components. In this case particularly, one of the biggest issues was the use
and dependence of the Axis26
classes for the web services request inside of an external
component, because of the final application size. Although, it didn’t exceed the size supported for
the server, the needed classes to represent an additional size for the application, which must be
considerated before use some tool in the development. This detail wasn’t taken in account in the
beginning of the process because of the lack of the use and performance of this kind of tools to
digital television.
6
http://axis.apache.org/axis2/java/core/

27
3. RELATED WORK
There have been apps about the use of: digital television, service adaptation and social network
together, but those works and apps have been treated in a separate way or using some of these
areas in an app, but there is not one that meets all the three topics. For this reason, we expose the
related works with this project independently.
In first place, taking the issue of user interaction-TV, we found two interesting works:
“Spontaneous interaction with audiovisual contents for personalized e-commerce over Digital
TV” [18] and “The Digital Television Multimedia Message Service Flows” [19], where is
important highlight their role in the context in which the applications are implemented, because
their focus is the user and not the machine or the tool itself. Some important developments are the
methods, platforms and tools used for each implementation, because these apps are so technical
and offer a framework of this point.
Related to the user – user interaction in digital television, we found some applications which are
focused in offering extra services for digital TV like “A T-Learning Platform based on Digital
Terrestrial Television” [20], “Spontaneous interaction with audiovisual contents for personalized
e-commerce over Digital TV”[18], “Exploiting digital TV users’ preferences in a tourism
recommender system based on semantic reasoning” [21] and “Proposed model of a digital video-
based home telecare system” [22]. These works show how the applications and tools that are
different to the TV content can be supported in this network. But, taking advantage of this
network and of the contents that are handled in it, the services can be enriched with such
information. Also, how the interaction is among users, the preferences of each one represent an
important difference to make the user feels comfortable and participant in the application.
In second place, we found some apps related to social networks and recommendation systems,
like a eco tourism system [23] and a recommendation system based on a social network [24],
where the main work there, is to use the social structure data to offer to users benefits according
to some user’s needs.
Other interesting applications are those which work as interaction facilitator. As mentioned
above, the center of the social net is the user, so, the relationships between them have a high
priority, because of indeed, each social activity is guided for the use of the social network [25].
4. CONCLUSIONS
• Using the interaction among users as base to offer services, allow to give better benefits
to users, because these services are going to be according to the information that we can
extract of the interactions, like the involved users, the type of correlation, the subject and
the related services, and to use this data for adapt services. In this way, it supports the
diffusion techniques and gives to users a better experience inside a system.
• The use of adaptation as tool to enrich services, allows to system to be prepared to face
different types of data, information, access devices, users and them preferences, so the
use environment for a client will be different for each one, compliance with specific
aspects of each like a personal preference or liking.
• The development of a digital TV application along with a social network and with
adapted services is a new proposal which offers benefits for developers, users and
providers. Also, it opens doors for the use, for research and for the implementation of
such tools to open the market and the possibilities for this kind of applications.

28
ACKNOWLEDGEMENTS
To the Pontificia Universidad Javeriana – Bogotá – Colombia, especially to the program of
Master in Information Systems and Computer Engineering and its research group SIDRe
(Information systems, distributed systems and networks).
REFERENCES
[1] CNTV, “CNTV.” [Online]. Available: http://www.cntv.org.co/cntv_bop/tdt/. [Accessed: 04-Mar-
2011].
[2] “TELEVISIÓN DIGITAL TERRESTRE,” WIKI TELEVISIÓN DIGITAL TERRESTRE. [Online].
Available: http://tdt.rtvc.gov.co/wiki/index.php/P%C3%A1gina_Principal. [Accessed: 06-Jun-2011].
[3] G. Adomavicius, Z. Huang, and A. Tuzhilin, “Personalization and Recommender Systems,” 2008
Tutorials in Operations Research: State-of-the-Art Decision-Making Tools in the Information-
Intensive Age, 2008.
[4] L. Li, Z. Yang, B. Wang, and M. Kitsuregawa, “Dynamic adaptation strategies for long-term and
short-term user profile to personalize search,” Advances in Data and Web Management, pp. 228–240,
2007.
[5] F. Qiu and J. Cho, “Automatic identification of user interest for personalized search,” in Proceedings
of the 15th international conference on World Wide Web, 2006, pp. 727–736.
[6] X. Shen, B. Tan, and C. X. Zhai, “Implicit user modeling for personalized search,” in Proceedings of
the 14th ACM international conference on Information and knowledge management, 2005, pp. 824–
831.
[7] M. Speretta and S. Gauch, “Personalized search based on user search histories,” in Web Intelligence,
2005. Proceedings. The 2005 IEEE/WIC/ACM International Conference on, 2005, pp. 622–628.
[8] Yuefeng Li and Ning Zhong, “Mining ontology for automatically acquiring Web user information
needs,” IEEE Transactions on Knowledge and Data Engineering, vol. 18, pp. 554–568, Apr. 2006.
[9] T. Robal and A. Kalja, “Applying user profile ontology for mining web site adaptation
recommendations,” in ADBIS, 2007, pp. 126–135.
[10] A. K. Dey and G. D. Abowd, “Towards a better understanding of context and context-awareness,” IN
HUC ’99: PROCEEDINGS OF THE 1ST INTERNATIONAL SYMPOSIUM ON HANDHELD
AND UBIQUITOUS COMPUTING, pp. 304–307, 1999.
[11] W3C, “Composite Capability/Preference Profiles (CC/PP): Structure and Vocabularies 1.0.” .
[12] J. Scott, Social network analysis: a handbook. SAGE, 2000.
[13] DVB project, “DVB,” DVB, 2003. [Online]. Available: http://www.dvb.org/. [Accessed: 05-Mar-
2011].
[14] Wikitel, “MHP.” [Online]. Available: http://es.wikitel.info/wiki/MHP. [Accessed: 03-May-2011].
[15] The MHP Knowledge Project (MHP-KDB), “The MHP-Guide.” .
[16] F. Ferrucho, “Taller Diseño de aplicaciones para TV Digital.” Linea de Tecnologías Virtuales -
Tecnoparque - SENA.
[17] Icareus Ltd, Icareus iTV Suite. Icareus Ltd.
[18] M. Lopeznores, M. Reylopez, J. Pazosarias, J. Garciaduque, Y. Blancofernandez, A. Gilsolla, R.
Diazredondo, A. Fernandezvilas, and M. Ramoscabrer, “Spontaneous interaction with audiovisual
contents for personalized e-commerce over Digital TV,” Expert Systems with Applications, vol. 36,
no. 3, pp. 4192–4197, Apr. 2009.
[19] L. Xing, J. Chun, and Y. Fan, “The Digital Television Multimedia Message Service Flows,” in
Networking and Digital Society, 2009. ICNDS ’09. International Conference on, 2009, vol. 2, pp.
267 –270.
[20] M. Baldi, A. De Santis, D. Falcone, E. Gambi, and S. Spinsante, “A T-Learning Platform based on
Digital Terrestrial Television,” in Software in Telecommunications and Computer Networks, 2006.
SoftCOM 2006. International Conference on, 292006-oct.1, pp. 347 –351.
[21] Y. Blanco-Fernández, “Exploiting digital TV users’ preferences in a tourism recommender system
based on semantic reasoning,” IEEE Transactions on Consumer Electronics, vol. 56, no. 2, pp. 904–
912, 2010.

29
[22] H. Lamminen, A. Lugmayr, S. Niiranen, and S. Kalli, “Proposed model of a digital video-based home
telecare system,” ℡EMEDICINE JOURNAL AND E-HEALTH, vol. 8, no. 4, pp. 387–394, WIN
2002.
[23] M. C. López Peredo, “Redes sociales en proyectos ecoturísticos,” 29-May-2011. [Online]. Available:
http://dialnet.unirioja.es/servlet/articulo?codigo=3233205. [Accessed: 29-May-2011].
[24] B. Fan, L. Liu, M. Li, and Y. Wu, “Knowledge Recommendation Based on Social Network Theory,”
in 2008 IEEE Symposium on Advanced Management of Information for Globalized Enterprises
(AMIGE), Tianjin, China, 2008, pp. 1–3.
[25] J. Song, W. Liu, and S. Chen, “Relation Grid: A Social Relationship Network Model,” in 2005 First
International Conference on Semantics, Knowledge and Grid, Guilin Guanxi, China, 2005, pp. 23–23.
Authors
Mery Yolima Uribe Rios MSc is a systems and computing engineer and magister, how
works and researches in networks, agents systems, digital television, security and grid
computing.
Rafael V. Páez PhD. works at Pontificia Universidad Javeriana and is one of the head
of in the research group SIDRe in distributed systems and networks area.

Recommendation System for Information Services Adapted, Over Terrestrial Digital Television

Related slideshows

More Related Content

Recommendation System for Information Services Adapted, Over Terrestrial Digital Television