Computational Intelligence - May 2016 - 38

GET http://esb.exactls.com/
COLLAGE/iccs/is?ticket=
{ticket}&query={term}
RETURN JSON array of objects
[
{"id": 0, "source": "twitter",
"term": "streaming report",
"data": ["Canadians
doubling time online with mobile devices,
video streaming: http://t.co/XCcRgmRj3a",],
"frequency": 0.06778420507907867},
{...}
]

CER uses linguistic descriptions of available and
required expertise and employs fuzzy computations
in order to recommend the right experts for
creative tasks.
3) Create a Team Retrieval (CER mashup): The following endpoint should be used to retrieve the creation of a team.
The first substep is to get the list of available skills and
subskills/tools.
GET http://www.htstats.com/collage/er/
{skills|tools}&token={token}
RETURN JSON object with the list of
skills or subskills

Once the available items are clear, the user can ask for translating his/her preferences to specific requirements:
POST http://www.htstats.com/collage/er/
translate&token={token}
RETURN JSON object with the explicit
requirements

The final step is to ask for the recommendation by passing
the requirements.
POST http://www.htstats.com/collage/er/
request&token={token}
RETURN JSON object with the list of
recommendations

All substeps should be run as many times as components of
the team we want.
4) Persona Retrieval (BrightSparks mashup): The following
endpoint should be used to retrieve one or all persona
information pertinent to the search parameter.
GET http://achernar.soi.city.ac.uk/
HallOfFame/HallOfFameService/Service1.
asmx?op=RetrieveAllPersonas
RETURN JSON object containing all current
personas with image and natural language
statements describing attributes of the
personas
GET http://achernar.soi.city.ac.uk/
HallOfFame/HallOfFameService/Service1.
asmx?op=RetrievePersonaAdvanced
RETURN JSON object about the persona

5) Search the Web for Inspirational Content (CRUISE mashup): The following endpoint should be used to invoke CRUISE to search
the Web or specific company sources for inspirational content:

38

IEEE ComputatIonal IntEllIgEnCE magazInE | may 2016

Note that the frequency count can be used by the user interface rendering code to size the scale of the fonts used to display
results, effectively creating a tag cloud effect (see Figure 1).
Results may be text-based, images, or videos. Also note that as
query terms for this service one can use the JSON object
retrieved using the BrightSparks service in the previous step so
that persona clues and characteristics to be used to trigger
CRUISE to search for inspirational content.
6) User Interface Development: The user interface of IdeaFocus has been developed using the iframe Java programming
paradigm. Iframes allow services to communicate with the
parent page through events: the sender (the CI service inside
the iframe) sends a string to the parent window via a JavaScript event such as the following:
window.parent.postMessage("my string",
"*");

Instead of using a simple string, we use a JSON string in
order to manage the passing of parameters:
var myEvent = {
event:"search-cruise",
term: "term to search"
};
window.parent.postMessage(JSON.
stringify(myEvent), "*");

The event in the above example is used to open CRUISE
with the "term to search" from BrightSparks. The parent window has to register an event handler to capture the events from
the COLLAGE applications in the iframe:
window.addEventListener('message',
function (e) {
var _event = JSON.parse(e.data);
if (_event.event == "search-cruise") {
$(window).trigger("search-cruise",
{term: _event.term});
}
});
http://esb.exactls.com/ http://www.t.co/XCcRgmRj3a" http://www.htstats.com/collage/er/ http://www.htstats.com/collage/er/ http://www.htstats.com/collage/er/ http://achernar.soi.city.ac.uk/ http://achernar.soi.city.ac.uk/
Table of Contents for the Digital Edition of Computational Intelligence - May 2016
