1731 research outputs found
Sort by
Experiments and modelling of vertically coupled Microresonators
No full textMicroresonators are fundamental building blocks in the growing field of integrated photonics and several resonator-based devices such as filters, switches and routers are currently used in common optical telecommunication networks. In order to exploit the peculiar features offered by integrated resonators, a complete and consistent comprehension of their physics and of the processes they can accommodate is needed. More specifically, coupling of light to and from a resonator represents a crucial point: a correct comprehension of the coupling dynamics, a proper model for the system and its validation through experimental procedure are all essential elements for a fruitful exploitation of the device. Among the different resonator-waveguide coupling schemes, the most widely used is the in-plane coupling and it consists of a waveguide placed near to a microresonator and laying on the same plane. However, an alternative approach is represented by the vertical coupling scheme, where the waveguide lays under the resonator edge. The peculiar position of the waveguide in this last configuration causes the device to show specific properties not present in other common coupling schemes: namely, a working range spanning from almost visible wavelengths (780nm) to the near IR domain (1600nm), the selective excitation of high order resonator radial modes and the possibility to fabricate wedge and free-standing resonators without any detrimental effect on the bus waveguide.
In order to fully exploit these and other features of the vertical coupling scheme, a detailed investigation has been carried out throughout this thesis. The waveguide-microresonator system has been studied at different levels, from the general coupling dynamics to more specific and peculiar phenomena. In particular, the basic model proposed for the vertical coupling has been extended to consider wavelength dependences and an experimental validation has been carried out consequently. The reactive coupling model, which describes the internal dynamics of a vertically coupled resonator in the case of multimodal operation, has been experimentally proven. A general model considering the presence counterpropagating modes has been theoretically proposed and experimentally investigated. Finally, the bistable behaviour generated by thermo-optic effect when a large amount of power circulates in the microresonator has been experimentally studied.
In order to better characterize the system response, a specific interferometric setup has been implemented. It consists in a Mach-Zehnder computer driven interferometer, whose peculiar characteristic is the ability to perform simultaneous pump and probe transmittance and phase measurements of any integrated photonic device provided with input and output ports. In this way, the information carried by the phase of the propagating optical signal is added to the one provided by its intensity and contributes to produce a more complete picture of the investigated system. In the case of microresonators this phase information becomes even more fundamental. Indeed, the phase response of a resonating structure is highly influenced by variations in the coupling strength, and the phase spectrum of a single resonance allows to clearly identify the resonator coupling regime for that specific resonance. This fact does not hold in the case of transmittance measurements, where single resonance spectrum carries information only on the total losses of the system. Finally, in order to exploit the combined information provided by this measurement procedure, a phasor plot representation is extensively used throughout the thesis work
An Online Peer-Assessment Methodology for Improved Student Engagement and Early Intervention
Get PDFStudent performance is commonly measured using summative assessment methods such as midterms and final exams as well as high-stakes testing. Although not as common, there are other methods of gauging student performance. Formative assessment is a continuous, student-oriented form of assessment, which focuses on helping students improve their performance through continuous engagement and constant measurement of progress. One assessment practice that has been in use for decades in such a manner is peer-assessment. This form of assessment relies on having students evaluate the works of their peers. The level of education in which peer-assessment is used may vary across practices. The research discussed here was conducted in a higher education setting. Despite its cross-domain adoption and longevity, peer-assessment has been a practice difficult to utilize in courses with a high number of students. This directly stems from the fact that it has been used in traditional classes, where assessment is usually carried out using pen and paper. In courses with hundreds of students, such manual forms of peer-assessment would require a significant amount of time to complete. They would also contribute much to both student and instructor load. Automated peer-assessment, on the other hand, has the advantage of reducing, if not eliminating, many of the issues relating to efficiency and effectiveness of the practice. Moreover, its potential to scale up easily makes it a promising platform for conducting large-scale experiments or replicating existing ones. The goal of this thesis is to examine how the potential of automated peer-assessment may be exploited to improve student engagement and to demonstrate how a well-designed peer-assessment methodology may help teachers identify at-risk students in a timely manner. A methodology is developed to demonstrate how online peer-assessment may elicit continuous student engagement. Data collected from a web-based implementation of this methodology are then used to construct several models that predict student performance and monitor progress, highlighting the role of peer-assessment as a tool of early intervention. The construction of open datasets from online peer-assessment data gathered from five undergraduate computer science courses is discussed.
Finally, a promising role of online peer-assessment in measuring levels of student proficiency and test item difficulty is demonstrated by applying a generic Item Response Theory model to the peer-assessment data
Strongly correlated quantum fluids and effective thermalization in non-Markovian driven-dissipative photonic systems
Get PDFCollective quantum phenomena are fascinating, as they repeatedly challenge our comprehension of nature and its underlying mechanisms. The qualification ``quantum'' can be attributed to a generic many-body system whenever the interference effects related to the underlying wave nature of its elementary constituents can not be neglected anymore, and a naive classical description in terms of interacting billiard balls fails to catch its most essential features. This interference phenomenon called ``quantum degeneracy'' which occurs at weak temperatures, leads to spectacular collective behaviours such as the celebrated Bose-Einstein Condensation (BEC) phase transition, where a macroscopic fraction of a bosonic system of particles collapses below a critical temperature T_c on a single-particle state. Quantum coherence, when combined with inter-particle interactions, gives rise to highly non-classical frictionless hydrodynamic behaviours such as superfluidity (SF) and superconductivity (SC).
Even more exotic quantum phases emerge in presence of important interactions as matter reaches a ``strongly correlated regime'' dominated by quantum fluctuations, where each particle is able to affect significantly the surrounding fluid: characteristic examples are the so-called Mott-Insulator (MI) quantum phase where particles are localized on a lattice due to a strong interaction-induced blockade, along with the Tonks-Girardeau (TG) gas where impenetrable bosons in one-dimension acquire effective fermionic statistics up to a unitary transformation, and the Fractional Quantum Hall (FQH) effect which occurs in presence of a gauge field, and features a special type of elementary excitation possessing a fractional charge and obeying to fractional statistics called `anyon'.
These quantum many-body effects were explored in a first place in systems well isolated from the external environment such as ultra-cold atomic gases or electrons in solid-state systems, within a physical context well described by ``equilibrium statistical mechanics''. Yet, over the last two decades a broad community has started investigating the possibility of stabilizing interacting quantum phases in novel nonlinear quantum optics architectures, where interacting photons have replaced their atomic and electronic counterpart. Thanks to their high level of controllability and flexibility, and the possibility of reaching the quantum degeneracy regime at exceptionally high temperatures, these platforms appear as extremely promising candidates for the ``quantum simulation'' of the most exotic many-body quantum problems: while the precursors experiments in semiconductor exciton-polariton already allow to reach the Bose-Einstein Condensation and superfluid regimes, novel platforms such as superconducting circuits, coupled cavity arrays or photons coupled to Rydberg EIT (Electromagnetically induced Transparency) atoms have entered the so-called `photon blockade' where photons behave as impenetrable particles, and open a encouraging pathway toward the future generation of strongly correlated phases with light.
A specificity of quantum optics devices is their intrinsic ``non-equilibrium'' nature: the interplay between the practically unavoidable radiative and non-radiative losses and the external drive needed to replenish the photon gas leads the many-body system toward a steady-state presenting important non-thermal features. One one hand, an overwhelmingly large quantity of novel quantum phenomena is expected in the non-equilibrium framework, as breaking the thermal equilibrium condition releases severe constraints on the state of a quantum system and on the nature of its surrounding environment. On the other hand, we do not benefit yet of an understanding of non-equilibrium statistical mechanics comparable with its well-established equilibrium counterpart, which relies on strong historical foundations. Understanding how to tame (and possibly exploit) non-equilibrium effects in order to stabilize interesting quantum phases in a controlled manner often reveals a hard challenge.
In that prospect, an important conceptual issue in the non-equilibrium physics of strongly interacting photons regards the possibility of stabilizing ``incompressible quantum phases'' such as the Mott-Insulator or Fractional Quantum Hall states, and more generally to stabilize the ground-state of a given particle-number conserving Hamiltonian, in a physical context where dissipative losses can not be neglected. While being able to quantum simulate those emblematic strongly correlated quantum phases in this novel experimental context would strongly benefit to the quantum optics community, gaining such a kind of flexibility would also contribute to fill an important bridge between the equilibrium and the non-equilibrium statistical physics of open quantum systems, allowing to access in a controlled manner a whole new phenomenology at the interface between the two theories.
In this thesis I address those questions, which I reformulate in the following manner:
-What are the conditions for the emergence of analogue equilibrium properties in open quantum systems in contact with a non-thermal environment ?
-In particular, is it possible to stabilize strongly correlated quantum phases with a perfectly defined particle number in driven-dissipative photonic platforms, in spite of environment-induced losses and heating effects ?
The structure of the thesis is the following.
[Chapter 1.] We give an overview of the physics of many-body photonic systems. As a first step we address the weakly interacting regime in the physical context of exciton-polaritons: after describing the microscopic aspects of typical experiments, we move to the discussion of non-equilibrium Bose-Einstein Condensation and the various mechanisms related to the emergence of thermal signatures at steady-state. The second part of this Chapter is dedicated to strongly interacting fluids. After drawing a quick overview of several experimental platforms presenting a good potential for the study of such physics in a near future, we discuss the relative performance of several schemes proposed in order to replenish the photonic population
[Chapter 2.] We investigate the potential of a non-Markovian pump scheme with a narrow bandpass (Lorentzian shaped) emission spectrum for the generation of strongly correlated states of light in a Bose-Hubbard lattice. Our proposal can be implemented by mean of embedded inverted two-level emitters with a strong incoherent pumping toward the excited state. Our study confirms in a single cavity the possibility of stabilizing photonic Fock states in a single configuration, and strongly localized n=1 Mott-Insulator states in a lattice with n=1 density. We show that a relatively moderate hopping is responsible for a depletion of the Mott-state, which then moves toward a delocalized state reminiscent of the superfluid regime. Finally, we proceed to a mean-field analysis of the phase diagram, and unveil a Mott-to-Superfluid transition characterized by a spontaneous breaking of the U(1) symmetry and incommensurate density. The results of this Chapter are based on the following publications:
- J. Lebreuilly, M. Wouters and I. Carusotto, ``Towards strongly correlated photons in arrays of dissipative nonlinear cavities under a frequency-dependent incoherent pumping'', C. R. Phys., 17(8), 836, 2016.
- A. Biella, F. Storme, J. Lebreuilly, D. Rossini, R. Fazio, I. Carusotto and C. Ciuti, ``Phase diagram of incoherently driven strongly correlated photonic lattice'', Phys. Rev. A, 96, 023839, 2017.
[Chapter 3.] In view of improving the performance of the scheme introduced in last chapter, and reproducing in particular the equilibrium zero temperature phenomenology in driven-dissipative photonic lattices, we develop a fully novel scheme based on the use of non-Markovian reservoirs with tailored broadband spectra which allows to mimick the effect of tunable chemical potential. Our proposal can be implemented by mean of a small number of emitters and absorbers and is accessible to current technologies.
We first analyse the case of a frequency-dependent emission with a square spectrum and confirm the possibility of stabilizing Mott insulator states with arbitrary integer density. Unlike the previous proposal the Mott state is robust against both losses and tunneling. A sharp transition toward a delocalized superfluid-like state can be induced by strong values of the tunneling or a change in the effective chemical potential. While an overall good agreement is found with the T=0 predictions, our analysis highlights small deviations from the equilibrium case in some parts of the parameters space, which are characterized by a non-vanishing entropy and the kinetic generation of doublon excitations. We finally consider an improved scheme involving additional frequency-dependent losses, and show in that case that the Hamiltonian ground-state is fully recovered for any choice of parameters. Our proposal, whose functionality relies on generic energy relaxation mechanisms and is not restricted to the Bose-Hubbard model, appears as a promising quantum simulator of zero temperature physics in photonic devices. The results of this Chapter are based on the following publication:
- J. Lebreuilly, A. Biella, F. Storme, D. Rossini, R. Fazio, C. Ciuti and I. Carusotto, ``Stabilizing strongly correlated photon fluids with non-Markovian reservoirs'', Phys. Rev. A 96, 033828 (2017).
[Chapter 4.] We adopt a broader perspective, and analyse the conditions for the emergence of analogous thermal properties in driven-dissipative quantum systems. We show that the impact of an equilibrated environment can be mimicked by several non-Markovian and non-equilibrated reservoirs. Chapter 2 already features a preliminary result in that direction, showing that in presence of a broad reservoir spectral density a given quantum system will evolve toward a Gibbs ensemble with an artificial chemical potential and temperature. In this chapter we develop a broader analysis focusing as a counterpart part on the exactly solvable model of a weakly interacting Bose Gas in the \acs{BEC} regime. Our formalism based on a quantum Langevin model, allows in particular to access both static and dynamical properties: remarkably, we demonstrate not only the presence of an equilibrium static signature, but also the validity of the fluctuation-dissipation theorem. While our results apply only for low-energy excitations for an arbitrary choice of reservoir spectral densities, we predict that a fine tuned choices of reservoirs mimicking the so-called Kennard Stepanov condition will lead to a full apparent equilibration. Such effect that we call ``pseudo-thermalization'' implies that under very specific conditions, an open quantum system can present all the properties of an equilibrated one in spite of the presence of an highly non equilibrated environment. The results of this Chapter are based on the following paper:
- J. Lebreuilly, A. Chiocchetta and I. Carusotto, ``Pseudo-thermalization in driven-dissipative non-Markovian open quantum systems'', arXiv:1710.09602 (submitted for publication)
Modelling of high-energy grinding processes
Get PDFThe comminution and tuning of several structural parameters of materials is often accomplished following a top-down route by high energy grinding. The reduced size of the particles constituting the end product and the incorporation of defects cause modified materials properties and increased solid-state chemical reactivity.
Among grinding devices, the planetary ball mill features high efficiency and versatility, being suitable for almost any kind of material, from metals and ceramics to organic compounds and pharmaceuticals. While its design and the working principle are rather simple, since grinding occurs by impacts between milling media (balls and jar) and mill charge, the characteristics of the end product strongly depend on a multitude of variables, determining balls trajectories and velocities and, in turn, the nature of impulsive forces exchanged during collisions. Numerical models can enourmously contribute to shed light on the process by providing the time evolution of kinematic and dynamic properties of milling media as well as quantities involved in contact events, permitting to understand the role of each milling variable and to design the characteristics of the end product. This Thesis chiefly proposes the implementation of a multibody dynamics model of the planetary ball milling process, its direct and indirect validation – respectively against movie collected in-operando and properties of the end product revealed by the analysis of X-ray powder diffraction data –, the evaluation of the effect of selected milling variables and the investigation of innovative solutions defining specific collisions features, obtained by the re-design of the jar shape. Some relevant case studies are also presented, namely the exfoliation of a bulk for the production of 2D nanostructured materials and the milling of the pharmaceutical compound Efavirenz, aimed at inducing structural and microstructural transformations enhancing dissolution properties
Pianificare percorsi di successo scolastico per studenti di origine migrante. Un mixed-method study nella scuola secondaria in Italia
No full textAssumendo in modo non neutrale il successo formativo degli studenti di origine migrante come parametro di valutazione dell'efficacia dei percorsi educativi, emergono pesanti criticità e profonde ineguaglianze nel sistema scolastico italiano, relative non solo al fenomeno degli abbandoni scolastici, ma anche a quello della canalizzazione formativa delle ragazze e dei ragazzi cosiddetti "post-migranti" nella scelta della scuola superiore. Dopo un iniziale approfondimento delle policies sottese alle politiche educative europee - con una decostruzione del concetto di insuccesso scolastico in una visione multi-prospettica radicata nella critical pedagogy - si presenta il complesso disegno di ricerca, un Mixed Methods Sequential Explanatory Design quan->QUAL con direzione teorica qualitativa, adottato per poter approfondire alcuni elementi emergenti dalla prima fase esplorativa di studio, che analizza le traiettorie formative di studenti non italiani attraverso sei anni di scuola. La fase intermedia sul prestigio scolastico ha permesso poi di individuare il campione di riferimento della seconda parte qualitativa, una Grounded Theory critica orientata alla giustizia sociale, che indaga i processi di successo e di insuccesso scolastico degli alunni/e di origine non italiana e che ha portato all'elaborazione di un modello teorico finale multidimensionale, testato attraverso un questionario strutturato. Il progetto triennale ha coinvolto – attraverso osservazioni etnografiche ricorsive, interviste focalizzate e in profondità e questionari - dirigenti, docenti, figure genitoriali, studenti e personale esterno di alcuni Istituti secondari di primo grado della provincia di Trento e Torino (Italia), Los Angeles (California) e Odense (Danimarca).
Considerando la valenza strategica degli esiti dell'indagine presentata, si forniscono in conclusione alcune linee d'azione per gli stakeholders coinvolti a livello educativo sui percorsi formativi degli studenti di origine migrante, al fine di focalizzare – con uno sguardo pedagogico – alcune linee di intervento efficaci per promuovere il successo scolastico e formativo delle ragazze e dei ragazzi più fragili
Better safe than sorry: new CRISPR/Cas9 tools for improved genome engineering
Get PDFCRISPR nucleases are efficient tools to edit cellular genomes in a variety of organisms. However, the in vivo application of this technology is still severely limited by unwanted genomic cleavages, that are further increased by long-term expression of the nuclease and can lead to unpredictable results. To address this limitation, we developed a yeast-based assay which allows to simultaneously evaluate the on- and off-target activity towards two engineered genomic targets in order to select optimized Streptococcus pyogenes Cas9 (SpCas9) variants. The screening of SpCas9 variants obtained by random mutagenesis of the Rec1-II domain allowed the identification of hits with increased on/off ratios. Through the combination of the identified mutations within a single variant we isolated the best performing nuclease, that we named evoCas9 (evolved Cas9). Side by side analyses with recently reported rationally designed variants demonstrated a significant improvement in fidelity of our evoCas9. In addition, to control Cas9 persistence into cells over time, we developed a Self-Limiting Cas9 circuitry for Enhanced Safety (SLiCES) which consists of an expression unit for SpCas9, a self-targeting sgRNA and a second sgRNA targeting a chosen genomic locus. This self-limiting circuit, by controlling Cas9 levels, results in increased genome editing specificity. For its in vivo utilization, we integrated SLiCES into a lentiviral delivery system (lentiSLiCES) via circuit inhibition to achieve viral particle production. Following its delivery into target cells, the lentiSLiCES circuit is switched on to edit the intended genomic locus while simultaneously promoting its own neutralization through SpCas9 inactivation. The two strategies here developed represent complementary approaches to address a major issue in the genome editing field. On one hand, by preserving target cells from residual nuclease activity, our hit and go SLiCES system increases the safety margins for genome engineering. On the other, if compared to published structure-guided protein engineering approaches, our in vivo screening increases the likelihood to identify the best combination of amino acid substitutions for the generation of novel, error-free SpCas9 and could represent a valid strategy to enhance the specificity of other RNA-guided nucleases
La giustizia di transizione in Italia. L'esperienza dopo la seconda guerra mondiale
No full textLa transizione italiana dal fascismo alla democrazia costituisce un paradigma della più recente categoria di giustizia di transizione. L’indagine studia il caso italiano secondo tale prospettiva specifica, intesa come gamma di processi e meccanismi associati con i tentativi di una società di affrontare un’eredità di passati abusi su larga scala. Vengono analizzate le soluzioni adottate in Italia, a partire dalla caduta del regime fascista, al fine di “fare i conti” con il passato. Lo studio fornisce una ricognizione prima di tutto sul piano del diritto penale, sia sostanziale che processuale, prendendo in esame sia gli interventi normativi che la prassi applicativa. Si passa poi all’esame di altri meccanismi, non solo sul piano amministrativo e civile, ma anche su quello politico e costituzionale, giungendo a riflettere su come il diritto penale si rapporti al processo di transizione nel suo insieme. In particolare, il lavoro analizza approfonditamente il rapporto fra transizione e potere di clemenza, partendo da un’analisi critica dell’amnistia Togliatti, vista come paradigma della transizione italiana. La prospettiva penalistica viene integrata con la comparazione giuridica e con la chiave di lettura della giustizia di transizione, che include una prospettiva storica, politica, sociologica e filosofica. Il lavoro legge il caso italiano come esempio di giustizia di transizione, lo colloca all’interno dello scenario comparato, verificando se si possa o meno ritenere esistente un modello italiano, cercando di offrire uno studio critico del processo transizionale nel suo insieme
Mondo, libertà, finitezza: Eugen Fink e la questione meontica dell'origine
No full textLa tesi ricostruisce l’evoluzione del pensiero di Eugen Fink, dalla meditazione fenomenologica degli anni Trenta alla filosofia cosmologica del dopoguerra, inaugurata nel 1946 dallo scritto Nietzsches Metaphysik des Spiels, mostrando come sia improprio parlare di una “svolta” cosmologica nel pensiero finkiano e scorgendo anzi nella Weltfrage, oltre che nel concetto della “meontica”, il filo conduttore che consente di cogliere la continuità della riflessione del filosofo: la questione cosmologica non entra in scena con il dopoguerra, ma è anzi ciò che motiva l’interesse di Fink per la fenomenologia di Husserl, costituendo il banco di prova di fronte al quale si deve provare la validità del metodo fenomenologico. Ulteriore scopo del lavoro è far emergere l’originalità della filosofia finkiana sia rispetto a Husserl, sia e soprattutto rispetto a Heidegger, sulla base dei manoscritti inediti conservati presso l’Universitätsarchiv della Albert-Ludwigs-Universität di Freiburg im Breisgau – di cui si propongono, tradotti, alcuni passi – nonché attraverso confronti più circoscritti con una serie di altri autori (tra i quali Marx, Adorno, Gehlen e, soprattutto, Patočka). Dopo un’introduzione volta a ricostruire lo stato attuale delle ricerche e le principali interpretazioni dell’opera finkiana, il primo capitolo illustra gli elementi di novità che la Sesta meditazione cartesiana introduce rispetto all’impostazione husserliana, individuando in tale scritto la prima effettiva presa di distanza di Fink dal maestro e scorgendo l’originalità del contributo finkiano soprattutto nel ruolo assegnato allo spettatore trascendentale, concepito come mero “esponente funzionale” della riflessione fenomenologica, operante uno sfondamento del piano ontologico della coscienza trascendentale. Dopo avere mostrato il debito di Fink nei confronti di Hegel per la concezione dialettica dell’Assoluto fenomenologico, si illustrano le differenze tra Fink e Husserl nel modo di intendere la “mondanizzazione secondaria”, nonché le ripercussioni della teoria della riduzione come de-umanizzazione (Entmenschung) sui compiti pedagogici della fenomenologia e in relazione alla possibilità di un’etica fenomenologica. Trait d’union tra primo e secondo capitolo è la critica del concetto husserliano di mondo. Il secondo capitolo muove da un confronto tra l’interpretazione heideggeriana di Hegel e quella finkiana, incentrata sul concetto cosmologico della “Vita”, per poi svolgere un confronto tra differenza ontologica e differenza cosmologica, principalmente alla luce del Colloquio sulla Dialettica e dell’Heraklit-Seminar, mostrando come la cosmologia finkiana, lungi dall’essere una mera integrazione dell’ontologia di Heidegger, implichi una diversa concezione dell’essere, realizzando una vera e propria “rivoluzione della comprensione ontologica”, che sfocia in una differente esplicazione della libertà, vista non più come corrispondenza alla chiamata dell’Essere all’interno dell’Ereignis, ma come l’auto-produzione (Selbstherstellung) dell’uomo. Il terzo capitolo mostra infine come la filosofia finkiana del dopoguerra sia di fatto l’articolazione cosmologica di quella “Lehre von der Freiheit” che costituiva il nucleo profondo della stessa fenomenologia husserliana: l’esame del gioco quale essenza positiva della libertà, il confronto con Patočka, e, non da ultimo, una lettura fenomenologica della nietzscheana “trasvalutazione dei valori”, sono i momenti teoretici che consentono di delineare i tratti di una originale e inedita “fenomenologia della libertà”, nella quale si esprime il medesimo ideale finkiano – inaugurato dalla radicalizzazione della riduzione fenomenologica come “riduzione tematica dell’idea di essere” nella Sesta meditazione cartesiana – della filosofia quale liberazione dall’ “irretimento mondano” (“Weltbefangenheit”)
Extraction and Exploitation of User Goals and Intentions for Querying and Recommendation
Get PDFUsers are often found in situations where they need to make selections from very large collections of items. These items may be digital artifacts e.g., web pages or forum posts, or digital representations of real world objects, e.g., products or people. There is a great deal of techniques for assisting users in making such selections. However, the plethora of systems and the size of the item collections makes the ability to provide the users with the items that really meet their standards in terms of interestingness and usefulness, a challenging task. We are dealing with the problem of providing items of interest to the users as response to explicit user requests or in the form of recommendations by exploiting a factor that has been poorly investigated so far in information systems: the goals for which items are intended, i.e., the goals for which items have been generated or produced; and the goals that may lead the user to “consume” them, i.e., the goals that s/he is willing to fulfill. The items may not be just items but interactions with items or actions that the user may be interested in performing. In this dissertation, we provide the required background and framework for exploiting goals in building better data managements systems.
Within this context, we study three different problems.
First, we are dealing with the problem of finding posts of interest (related posts) given a post-query in forums within user communities. Forum posts consist of segments each one serving a different goal that the author had in mind to communicate to the reader through the text. Therefore, plain content comparisons often fail to retrieve posts of interest, or they retrieve posts that despite the similar content are not related to the post-query. Instead, we have developed a goal-aware matching approach that uses content similarity over intention-based segmentations, i.e., over segments that are intended for different communication goals to perform more effective comparisons. Second, we are dealing with the goal-aware recommendation problem.
This problem, opposed to the post matching mechanism to which we have referred earlier does not consider domain specific characteristics; thus it can be applied to any domain. The goal-aware mechanisms we have developed handle the diverse goals that the user can fulfill by first recognizing the intended user goals, deciding the priorities among them, and by quantifying the benefit of each item.
Last but not least, we are dealing with the problem of building a goal implementation set from texts where users describe how they managed to fulfill certain goals in their real life. We have applied our technique on textual descriptions from a goal-setting site. For each solution we have designed, implemented and extensively evaluated models, algorithms and techniques that deal with all the individual tasks that are required for a goal-aware approach: the identification and extraction of goal-related information in the examined data sources, the modeling of the derived information, the matching of the user's request or previous activity to the goal model elements, and finally the exploitation of this matching into the forming of the system's response. The goal-aware techniques have been found to retrieve items that would not have been considered by the traditional techniques giving to the user a different and more complete view of the item collection. Moreover, the scalability of the techniques and the efficient structures and indexes that we use to store and retrieve the items alongside the goal-related data allows us to meet the requirements of modern online systems
Predictive Modeling of Human Behavior: Supervised Learning from Telecom Metadata
No full textBig data, specifically Telecom Metadata, opens new opportunities for human behavior understanding, applying machine learning and big data processing computational methods combined with interdisciplinary knowledge of human behavior. In this thesis new methods are developed for human behavior predictive modeling based on anonymized telecom metadata on individual level and on large scale group level, which were studied during research projects held in 2012-2016 in collaboration with Telecom Italia, Telefonica Research, MIT Media Lab and University of Trento. It is shown that human dynamics patterns could be reliably recognized based on human behavior metrics derived from the mobile phone and cellular network activity (call log, sms log, bluetooth interactions, internet consumption).
On individual level the results are validated on use cases of detecting daily stress and estimating subjective happiness. An original approach is introduced for feature extraction, selection, recognition model training and validation. Experimental results based on ensemble stochastic classification and regression tree models are discussed. On large group level, following big data for social good challenges, the problem of crime hotspot prediction is formulated and solved. In the proposed approach we use demographic information along with human mobility characteristics as derived from anonymized and aggregated mobile network data. The models, built on and evaluated against real crime data from London, obtain accuracy of almost 70% when classifying whether a specific area in the city will be a crime hotspot or not in the following month. Electric energy consumption patterns are correlated with human behavior patterns in highly nonlinear way. Second large scale group behavior prediction result is formulated as predicting next week energy consumption based on human dynamics analysis derived out of the anonymized and aggregated telecom data, processed from GSM network call detail records (CDRs). The proposed solution could act on energy producers/distributors as an essential aid to smart meters data for making better decisions in reducing total primary energy consumption by limiting energy production when the demand is not predicted, reducing energy distribution costs by efficient buy-side planning in time and providing insights for peak load planning in geographic space. All the studied experimental results combine the introduced methodology, which is efficient to implement for most of multimedia and real-time applications due to highly reduced low-dimensional feature space and reduced machine learning pipelines. Also the indicators which have strong predictive power are discussed opening new horizons for computational social science studies