Call for applications

Call for applications to work on the project named CLIPS

Art. 1 - Description of the activities to be performed by CRAT

CRAT has concluded a consulting agreement with Magna Gracia Interdepartmental Centre to collaborate in the project “Cloud-based Platform for Ommersive adaptive video Streaming – CLIPS” (Piano Opreativo Nazionale “Imprese e competitività” 20014/2020 FESR).

In particular, CRAT must perform the following activities:

 

Industrial Research

AR1: State of the art on Machine Learning algorithms

The aim of this activity is to study the state of the art relating to Data Mining and Machine Learning algorithms for User Profiling, QoE Evaluation and QoE Control modules. As regards the User Profiling module, the adaptability to the project of possible unsupervised learning algorithms will be studied, with particular reference to partitioning clustering algorithms such as k-means in the presence of large quantities of heterogeneous data; particular attention will be given to the big data aspects, strictly connected with the volume and heterogeneity of the data, as well as the efficiency and reliability of the algorithms to be used in an application in which scalability is an element of fundamental importance. As regards the QoE Evaluation module, the methods existing in the literature will be studied to evaluate the subjective quality perceived by users who use a video and, in particular, how this quality depends on appropriate parameters related to the user profile (identified by the module User Profiling), as well as by appropriate KPIs describing the objective performance of the video. Finally, as regards QoE Control, the adaptability of possible Reinforcement Learning algorithms (for example, Q-Learning, Sarsa, etc.) will be studied with particular attention to the problem of scalability.

AR2: Design of the User Profiling software component and of the Monitored Information Database

In this activity we will design the architecture of the Monitored Information Database and, in particular, the structure of the records of this database (Session Records), each of which will contain all the information deemed useful relating to a video session ended or in progress.

Furthermore, in this activity the machine learning and pattern recognition algorithms that will perform the non-real-time analysis of the data stored in the Monitored Information Database will be identified and implemented. This analysis aims to infer (i) the number of different User Profiles, (ii) the association between the users and the identified User Profiles, (iii) for each User Profile the parameters that characterize the QoE perceived by the users belonging to the User Profile in question.

AR3: Design of the QoE Evaluation software component

In this activity we will design an appropriate set of QoE Utility Functions customized with respect to the user profile, capable of evaluating, in real time, the QoE perceived subjectively by each individual user involved in a video session.

The aforementioned QoE Utility Functions will be characterized by the set of parameters associated with the User Profiles; their independent variables will consist of appropriate Key Performance Indicators (KPI) describing the objective performance of the video session for which the Perceived QoE is to be assessed. In this regard, the aforementioned KPIs will also be processed in this activity starting from the data stored in the Monitored Information Database.

AR4: Design of the QoE Control software component

In this activity the QoE Control functions will be designed which will deduce the reference performance parameters, customized for each single session, which the video and cloud resources control algorithms should try to obtain (or better to "chase") to maximize the degree of satisfaction from each user for each individual session (i.e. subjectively perceived QoE). The aforementioned performance benchmarks will be deduced by trying to infer, through appropriate learning algorithms, the cause-effect link between these parameters and the QoE perceived subjectively by users.

Experimental Development

AS1: Implementation of the User Profiling software component and of the Monitored Information Database

In this activity, the software related to the User Profiling component described in the activity "AR2: Design of the User Profiling software component and of the Monitored Information Database" will be developed and the related simulations will be developed to verify performance in complex scenarios, in line with the reference scenarios of the project. Furthermore, the Monitored Information Database described in the AR2 activity will be implemented and interfaced with the aforementioned User Profiling component.

AS2: Implementation of the QoE Evaluation software component

In this activity, the software related to the QoE Evaluation component described in the AR3 activity "Design of the QoE Evaluation software component" will be developed and the related simulations will be developed to verify performance in complex scenarios. Furthermore, this component will be integrated with the User Profiling component implemented in the AS1 activity "Implementation of the User Profiling software component and the Monitored Information Database".

AS3: Implementation of the QoE Control software component

In this activity, the software related to the QoE Control component described in activity "A5.4 Design of the QoE Control software component" will be developed and the related simulations will be developed for performance verification in complex scenarios. We will also integrate this component with that of QoE Evaluation implemented in AS2 activity "Implementation of the QoE Evaluation software component".

Art. 2 - How to submit an application

With this call we invite the candidates who have the skills needed to perform some of the activities relevant to the WPs mentioned in Art. 1 to send an email application to the address  This email address is being protected from spambots. You need JavaScript enabled to view it. , with subject “Application for working in CLIPS”, attaching (i) a signed curriculum vitae in .pdf format, (ii) a front/back copy of a) a valid identity card b) the fiscal code. Within the same terms, the application can be hand carried to the CRAT office, located in via Giovanni Nicotera 29, 00195, Rome (IT), Tuesday-Thursday 10:00-12:00.

The applications may be submitted at any time within the duration of the project (see Art. 1). The application is valid for the whole project duration. Each candidate is entitled to send at any time and in the way previously described any addendum to the curriculum.

Applications will be considered by the Selection Committee described in art. 3, according to the timetable specified in Art. 4, following the criteria specified in Art. 5.

 

Art. 3 - Composition of the Selection Committee

The Selection Committee will be composed of the Scientific Responsible(s) of the project and the members of the CRAT Scientific Committee more competent on the project's theme.

The selection committee will be appointed by the President of CRAT with the unanimous approval of the CRAT Board of Directors. The latter will supervise the selection process and will guarantee the absence of conflicts of interest.

 

 Art. 4 - Selection Committee timetable

The Selection Committee will hold meetings thrice a year, in the first half of February (winter session) in the first half of April (Spring session) and in the first half of September (Autumn session) , in order to assess the received applications aiming at identifying the new collaboration contracts to be awarded. In the each session the applications received until February 10, April 10 and September 10 will be screened.

 

Art. 5 - Selection Committee operations

Within each session, the Selection Committee, on the basis of the applications submitted within the deadlines set out in Art. 4, will update a short list including the eligible candidates, i.e. the candidates which are eligible to carry out all or part of one or more of the activities referred to in Art. 1.

An eligible candidate will remain in the short list for the whole duration of the project. The short list of eligible candidates may be requested by every candidate, at any time, by writing an email to the   address mentioned in art. 2. The candidates which will be awarded with a contract (hereinafter referred to as contractors) will be chosen within the short list in question. Nevertheless, the inclusion of a candidate in the short list does not imply any right on her/his part to get contract assignments.

The contractor selection, the duration of the contract (generally not less than 6 months), the type of contract, the hourly rate and the number of hours to be worked under the contract, will be identified by the Selection Committee on the basis of the following criteria:

-   Professional skills and experience of the candidate inferred from the curriculum attached to the application and, possibly, from an interview with the candidate; in particular, the following points will be taken into account: the academic rank (professor, researcher, post-doc, doctoral student, graduate student, ...), the merits and qualifications acquired in the framework of applied research, the organizational skills and the ability to work in a team;

- "best matching" between the eligible candidates, the man-months foreseen by the contract in the various activities (see Art. 1);

-   the principle of "best value for money";

-   equal opportunities for men and women;

- all the remaining criteria laid down in Art. 32 and 33 of the contract with the European Commission as detailed in the Annotated Model Grant Agreement and in the documents referred to herein.

 

After each session, the Selection Committee draws up a report which shall include, on the basis of these criteria, the candidates selected for the award of contracts, as well as the duration of the contract, the type of contract, the hourly rate and the number of hours to be worked under the contract.

A copy of this report can be requested by the applicants upon the submission of a written request addressed to CRAT, via Giovanni Nicotera 29, 00195, Rome (IT), by registered mail with return receipt, no later than January 15 for the winter session and July 15 for the summer session. Applicants may submit any counter respectively by January 31 (winter session) and by July 31 (summer session); the Selection Committee must give adequate response within 15 days.

 

Art. 6 - Contract actuation

The CRAT President, on the basis of the Selection Committee report, will prepare by the end of December (for the winter session), or by the end of June (for the winter session) the contracts and send them to the selected contractors via e-mail. These latter, within one week, should send to the President of CRAT a paper copy of the contract signed for acceptance.

The contractors will work under the CRAT instructions on the CRAT premises. They will be free to determine breaks, rest periods and, in general, working times; nevertheless, they will be paid on the basis of the hours actually worked on the project (which will be clearly reported on the contract); such hours must be monthly reported on proper time sheets, subject to the approval of a supervisor. The results of the work performed by the contractor will belong to CRAT.

 

Art. 7 - Protection of Privacy

Pursuant to Legislative Decree no. 196/2003 and the EU General Data Protection Regulation, GDPR, the personal data will be used exclusively for the purposes of managing the selection and, later, for purposes related to the conduct of the contract, in accordance with current rules on transparency and privacy protection.