ArchivIA Università degli Studi di Catania
 

ArchivIA - Archivio istituzionale dell'Universita' di Catania >
Tesi >
Tesi di dottorato >
Area 01 - Scienze matematiche e informatiche >

Please use this identifier to cite or link to this item: http://hdl.handle.net/10761/3974

Issue Date: 8-Feb-2016
Authors: Napoli, Christian
Title: A-I: Artificial intelligence
Abstract: In this thesis we proposed new neural architectures and information theory approaches. By means of wavelet analysis, neural networks, and the results of our own creations, namely the wavelet recurrent neural networks and the radial basis probabilistic neural networks,we tried to better understand, model and cope with the human behavior itself. The first idea was to model the workers of a crowdsourcing project as nodes on a cloud-computing system, we also hope to have exceeded the limits of such a definition. We hope to have opened a door on new possibilities to model the behavior of socially interconnected groups of people cooperating for the execution of a common task. We showed how it is possible to use the Wavelet Recurrent Neural Networks to model a quite complex thing such as the availability of resources on an online service or a computational cloud, then we showed that, similarly, the availability of crowd workers can be modeled, as well as the execution time of tasks performed by crowd workers. Doing that we created a tool to tamper with the timeline, hence allowing us to obtain predictions regarding the status of the crowd in terms of available workers and executed workflows. Moreover, with our inanimate reasoner based on the developed Radial Basis Probabilistic Neural Networks, firstly applied to social networks, then applied to living companies, we also understood how to model and manage cooperative networks in terms of workgroups creation and optimization. We have done that by automatically interpreting worker profiles, then automatically extrapolating and interpreting the relevant information among hundreds of features for each worker in order to create workgroups based on their skills, professional attitudes, experience, etc. Finally, also thanks to the suggestions of prof. Michael Bernstein of the Stanford University, we simply proposed to connect the developed automata. We made use of artificial intelligence to model the availability of human resources, but then we had to use a second level of artificial intelligence in order to model human workgroups and skills, finally we used a third level of artificial intelligence to model workflows executed by the said human resources once organized in groups and levels according to their experiences. In our best intentions, such a three level artificial intelligence could address the limits that, until now, have refrained the crowds from growing up as companies, with a well recognizable pyramidal structure, in order to reward experience, skill and professionalism of their workers. We cannot frankly say whether our work will really contribute or not to the so called "crowdsourcing revolution", but we hope at least to have shedded some light on the agreeable possibilities that are yet to come.
Appears in Collections:Area 01 - Scienze matematiche e informatiche

Files in This Item:

File Description SizeFormatVisibility
NPLCRS85M14C351D-main-a.pdfmain-a6,98 MBAdobe PDFView/Open


Items in ArchivIA are protected by copyright, with all rights reserved, unless otherwise indicated.


Share this record
Del.icio.us

Citeulike

Connotea

Facebook

Stumble it!

reddit


 

  Browser supportati Firefox 3+, Internet Explorer 7+, Google Chrome, Safari

ICT Support, development & maintenance are provided by the AePIC team @ CILEA. Powered on DSpace Software.