<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta />
    <article-meta>
      <title-group>
        <article-title>Questioning Mashups</article-title>
      </title-group>
      <contrib-group>
        <aff id="aff0">
          <label>0</label>
          <institution>Rheinstr.</institution>
          <addr-line>1A 30519 Hannover</addr-line>
          <country country="DE">Germany</country>
        </aff>
      </contrib-group>
      <abstract>
        <p>Mashups have been de ned in earlier development stages of the web. Their main motivation is to build new services from existing resources. This principle is still productive. Assumptions on what mashups take in, which interfaces and services they provide and where they are networked need an overhaul. Focusing on a reuse of resources exploiting all their a ordances while opening wider ranges for all other features might bring the concept up to observable applications of today.</p>
      </abstract>
      <kwd-group>
        <kwd>Mashup</kwd>
        <kwd>Properties</kwd>
        <kwd>Application</kwd>
        <kwd>Network</kwd>
        <kwd>Interface</kwd>
        <kwd>Media</kwd>
        <kwd>E-Learning</kwd>
        <kwd>Bionics</kwd>
        <kwd>Automotives</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>-</title>
      <p>{ Excluded from the mashup league are applications that draw on one source
only and that are de nitively without any contact to a network / the web.
The basic mashup de nition needs an overhaul. It may be outdated. Maybe
more elaborated mixes of input media and interfaces overgrew the initial mashup
concept, while the driving force of the approach became more and more popular:
the reuse of existing resources for new purposes and services, exploiting all their
a ordances.
2</p>
    </sec>
    <sec id="sec-2">
      <title>Mashup making</title>
      <p>Before the advent of the web there were no (web) mashups, only their earlier
namesakes in music. Indeed mashups depend on the plenty of APIs and data
resources that the web ecosphere o ers { all items owning a URL can participate
in a mashup.</p>
      <p>First and foremost, a good mashup idea is needed for a good mashup. A
welldesigned mashup may be easy to understand. However it may be hard to
implement, given the divergent standards and aims of software and data in the web.
The task is to integrate resources that were developed for all sorts of things,
but certainly not for an easy reuse in some mashup. For good reasons, mashup
developers are strong partisans of web standards.</p>
      <p>
        Clashes and incompatibilities are lurking on many places. Figure 1 displays a
mashup construction scheme organized in layers. It illustrates the many entities
that must be integrated (more detail on mashup construction in [
        <xref ref-type="bibr" rid="ref6">6</xref>
        ], [
        <xref ref-type="bibr" rid="ref2">2</xref>
        ]).
Please note that user roles are missing in the gure. But they are important.
Mashups serve strong user roles and exible web services. Users putting together
their mashups with the help of mashup editors can combine resources so that
they serve their individual needs.
3
      </p>
    </sec>
    <sec id="sec-3">
      <title>Good old pioneering mashups</title>
      <p>From the start-up period till today geo mashups are very popular. They appear
as maps showing where local businesses and other establishments are located,
often with their photos and address data or with widgets, depending on the
purpose of the map. Today geo mashups cover most of the world. On gure 2 you
see an example: the Google map of my home. The classical Chicago Crime Map3
displays thefts, shootings and so on. Street views and car navigation services are
popular for good reasons: they really help users in everyday life.</p>
    </sec>
    <sec id="sec-4">
      <title>Probing some mixed-input mixed-interface items</title>
      <p>The web keeps accumulating data, media and additional resource types, such
as software services (SaaS { Software as a Service) or speci c networks such as
3 http://spotcrime.com/il/chicago
LOD (Linked Open Data). Preferring them to own items often is a question of
mere practicality. In other cases third-party resources are an enabling condition
of an application.</p>
      <p>Furthermore the web is entering deeper and deeper into our everyday life. Mashups
and related mixed-mode systems do as everybody does - they populate fresh
virgin soil of the web. They go mobile, they become ubiquitous, they are reaching
new user groups and they conquer new substrates, for instance with sensors
sitting on wearables like a jacket or an iWatch.</p>
      <p>Let us inspect some application areas for probing mashups and more in general
applications integrating several input resources and having possibly a few
interfaces. Moreover we have to be attentive to the networks that may come up. We
assume that the choice of domains does not really matter - one might obtain the
same overall impression from other elds.
4.1</p>
      <sec id="sec-4-1">
        <title>Audio and tactile</title>
        <p>SubPac4 accepts music from a mixer or a sound le. Music can be seen as a
couple of tracks, or of frequencies, as an ensemble of voice, violin etc., so that
one can consider it as a set of inputs as well. The SubPac service isolates the
low frequency bass and renders it to users as a rhythm felt by the touch sense as
in a discotheque under full sound. The physical interface is a sort of cushion or
backpack. With it the listener can hear and feel the music. By converting sound
to touch SubPac is rendering its users an interesting service. It delivers a tactile
interface.
4.2</p>
      </sec>
      <sec id="sec-4-2">
        <title>Virtual patients in e-learning</title>
        <p>
          When a physician examines a patient she collects all sorts of details, from the
prehistory of the patient to recent items, e.g. MRT (magnetic resonance
tomography) images or lab ndings. Figure 3 gives you an idea. The doctor follows
pre-established rules. The situation is complex enough for training physicians
(as pilots) in a ( ight) simulator environment. A popular US o er is MedSims5.
InMedea6 distributes a simulator developed in Germany. There are more
approaches on the market [
          <xref ref-type="bibr" rid="ref1">1</xref>
          ].
        </p>
        <p>Imagine connecting all the data for the virtual test patients, in addition all
the items the physician will provide while dealing with the patient, and the
allembracing clinical setting. As the simulator is dedicated to e-learning / physician
training, there may be a tutor whom the doctor can ask and who assesses the
work of the trainee.</p>
        <p>Authoring a virtual patient test case implies the coordination of many
information items. Intelligent integration and display of media clearly is a main task,
but probably the author will perform well without even knowing the concept of
mashups.
4 http://www.thesubpac.com
5 http://www.medsims.com
6 http://www.inmedea-simulator.net/med/scene/entry?</p>
      </sec>
      <sec id="sec-4-3">
        <title>4.3 Intelligent cars</title>
        <p>Internet services are entering the cars of almost all manufacturers. Apple
Car Play inside a Mercedes car7 of today ( gure 4) displays mobile multimedia
information from an iPhone on the car network, with Siri speech processing
included. The VW intelligent car vision8 on gure 5 receives and processes an
impressive quantity of information channels. Its service is to be autonomous
driving.
7 https://www.youtube.com/watch?v=SMX741OT0gg&amp;feature=youtu.be
8 https://www.itu.int/dms\_pub/itu-t/oth/06/1B/T061B0000020056PDFE.pdf</p>
        <p>Nobody doubts that humans and things are legitimate participants of mashups
and the like. Why should parts of humans and things remain out while humans
and things are in? Please consider an arti cial limb9 on gure 6: The service
is that a person can walk. Several stimuli from the body network come in via
9 https://www.youtube.com/watch?v=CDsNZJTWw0w
di erent pathways. Some software with cognitive foundation is at work and a
wireless link as well. If we want highly productive resource mixes - should they
not include exoskeletons and other applications we did not imagine before? And
by the way, where is the interface?
5</p>
      </sec>
    </sec>
    <sec id="sec-5">
      <title>Conclusion</title>
      <p>The skilled reuse of resources is a main impulse of the mashup concept, often
seen together with adaptive services and interfaces for users. It is widely adopted
today. As everyday things, networked things / resources have a ordances that
lie in the eyes of the beholder. They can be applied to unforeseen purposes.
Networking enables this reuse.</p>
      <p>The web made all sorts of resources available for further use: APIs, databases,
ontologies, software services, cloud storage and so on. This was a signal for
setting up mashups. Other networks, e.g. more local or restricted ones may o er
to mixed-resources applications all items they need, too.</p>
      <p>The inputs and interfaces of mixed-resources applications play on all available
channels: graphic, audio, tactile, interactive, operative, mixed, etc. One may even
ask whether a palpable interface must exist at all and which functions it should
keep. It may reduce to a start button in an intelligent car.</p>
      <p>The conclusion is evident: Let us replace the mashup concept which was good in
the beginnings of the web with a new version that concentrates on the intelligent
remix of resources on networks. No preset limits { let us challenge researchers
to explore how far they can go.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          1.
          <string-name>
            <given-names>Delgado</given-names>
            <surname>Lopez Paula</surname>
          </string-name>
          (
          <year>2013</year>
          ).
          <source>Virtual Patient</source>
          .
          <fpage>1</fpage>
          -6 in Martinez et al ed.
          <source>AITA</source>
          <year>2013</year>
          , Workshop on Ambient Intelligence for Telemedicine and Automotive
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          2.
          <string-name>
            <surname>Endres-Niggemeyer Brigitte</surname>
          </string-name>
          (
          <year>2013</year>
          ).
          <source>Semantic Mashups. Intelligent Reuse of Web Resources</source>
          . Springer Heidelberg
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          3.
          <string-name>
            <given-names>Fichtner</given-names>
            <surname>Darlene</surname>
          </string-name>
          (
          <year>2009</year>
          )
          <article-title>What Is a Mashup? 3-</article-title>
          17 in Engard Nicole C. ed. Library Mashups:
          <article-title>Exploring New Ways to Deliver Library Data</article-title>
          .
          <source>BooksInfoToday</source>
          <year>2009</year>
          . http://books.infotoday.com/books/Engard/Engard-Sample-Chapter.pdf
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          4.
          <string-name>
            <given-names>Malki</given-names>
            <surname>Abdelhamid</surname>
          </string-name>
          &amp;
          <string-name>
            <surname>Benslimane Sidi Mohammed</surname>
          </string-name>
          (
          <year>2012</year>
          ).
          <article-title>Building Semantic Mashup</article-title>
          .
          <source>Proceedings ICWIT</source>
          <year>2012</year>
          ,
          <volume>40</volume>
          -
          <fpage>49</fpage>
          . http://ceur-ws.
          <source>org/</source>
          Vol-
          <volume>867</volume>
          /Paper5. pdf
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          5.
          <string-name>
            <given-names>Martinez</given-names>
            <surname>Natividad</surname>
          </string-name>
          , Seepold Ralf &amp; Ortega Juan Antonio ed.
          <source>AITA</source>
          <year>2013</year>
          , Workshop on Ambient Intelligence for Telemedicine and Automotive / ISBN:
          <fpage>978</fpage>
          -
          <lpage>84</lpage>
          - 697-0147-8 http://madeirasic.us.es/aita2013/wp-content/uploads/2014/04/ actas\_web.pdf
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          6.
          <string-name>
            <surname>Yu</surname>
            <given-names>Jin</given-names>
          </string-name>
          , Benatallah Boualem, Casati Fabio &amp; Daniel
          <string-name>
            <surname>Florian</surname>
          </string-name>
          (
          <year>2008</year>
          )
          <article-title>Understanding Mashup Development</article-title>
          .
          <source>IEEE Internet Computing Sept/Oct</source>
          <year>2008</year>
          ,
          <fpage>44</fpage>
          -
          <lpage>52</lpage>
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>