PROJECT TITLE – Social media and digital interaction intelligence
ACRONYM- SoMeDi
Project’s identification number: 15001
Project partners:
- Spain, HI Iberia Ingeniería y Proyectos, Experis ManpowerGroup, S.L.U., Taiger
- Romania, BEIA Consult International, Siveco Romania S.A.
- Turkey, EVAM, Turkgen, Turkcell Teknoloji
Project leader: Raúl Santos De La Cámara [HI Iberia Ingenieria y Proyectos]
Contracting Authority in Romania – UEFISCDI
Deposit code: PN-III-P3-3.5-EUK-2016-0013
Contract number 90 din 01/09/2016
The total value of the budget: 2.099.865,00 Lei
The total value of the contract: 3.843.650,00 Lei
The total value of the co-founding: 1.743.785,00 Lei
Contract duration – 36 months
Project duration – 72 months More information on ITEA3_SOMEDI
SoMeDi Project profile leaflet
The amount of digital interaction data has soared along with the digitalization of business processes and private communication since the advent of the Internet. Cisco indicates that the annual global IP (Internet Protocol) traffic will pass the zettabyte (1000 exabytes) threshold by the end of 2016.
These ubiquitous engagements produce an almost unfathomable amount of interaction traces that provide information about preferences, interests, and intentions. At the same time advances in Artificial Intelligence, Data Analysis and Machine Learning to provide the opportunity to extract and use this information to better serve and engage our users and audiences. The ever-growing functionalities in our software, the breadth of our media asset libraries and the amount of potential messages stress our users with information overload, but with the help of Digital Interaction Intelligence, improvements like personalized offerings and tailored engagement can be applied to help users and increase the value of our systems.
SoMeDi will unlock the implicit value in digital interaction data by turning it into Digital Interaction Intelligence that can drive and support various business processes and use situations. Our consortium has attracted both leading technology experts and strong use case owners from B2C and B2B platforms in SaaS, Digital security, Digital Entertainment, and Gaming, as well as strong European Brands with vital interests in Competitive Intelligence and Customer Engagement in highly competitive global markets. Demonstrators, as well as experimentation scenarios across different vertical applications, ensure that the advances in research and technology can be used from all angles under real industrial constraints and lead to relevant economic valorization.
SoMeDi’s goal is to unlock the value hidden in the digital content and traces of human (inter) actions through the application of advanced artificial intelligence and machine learning techniques. To reach this goal, SoMeDi will:
- To develop novel methods and tools on top of general machine intelligence toolboxes, frameworks and services to efficiently analyze digital interaction data including social media. These include (1.1) methods to produce improved sentiment analysis and opinion mining to increase perception of the user’s attitude towards topics and concepts at aspect level, (1.2) methods that analyze user activities and online-phenomenon, such as detect communities and spread of messages in social networks, and (1.3) methods for providing decision-support of the analyzed data for different purposes of use.
- Connect the value mined through these methods with a company’s production and business processes, to improve offers and personalize service or content configurations.
- Explore how to use social media actively and efficiently growth hacking and marketing both in B2C and B2B markets, developing a methodology for effectively entering and acting in social media and use it for growth hacking, customer support, and marketing in connection to both business users and consumers.
Produce a guidebook introducing a practical Digital Interaction and Social Media Intelligence methodology with techniques, illustrated by the project’s use cases.
Phase I – Analysis of the context and users’ requirements
Phase I consists of the following activities:
- Activity 1.1 Analysis of the context and State Of The Art (correlated with T1.1)
- Activity 1.2 Defining the Digital Interaction Solutions Requirements and Specifications (correlated with T1.2)
Phase description
Phase I of the SoMeDi project ran from 01.09.2016 to 30.12.2016 and aimed at analyzing the context and the requirements of the users. The objectives of this step are related to T1.1 of the SoMeDi project, which focuses on the analysis of the current State Of The Art, as well as the T1.2 which refers to defining the requirements and specifications of the solution for digital interactions.
Results
Phase I of the project produced one market study and one research report.
The results achieved within phase I have been disseminated in the following event: participation in IMWorld 2016.
Phase II – Definition of the techniques for Digital Interaction
Phase II is comprised of the following activities:
- Activity 2.1 Analysis of Digital Interaction Solutions
- Activity 2.2 Current tools for Digital Interactions
- Activity 2.3 Techniques for Modeling and Searching in Metadata
Phase description
The 2nd Phase of the SoMeDi project ran from 01.01.2017 to 30.05.2017 and focused on the State-Of-The-Art (SoTA) technology and tools currently being developed in the field of Intelligent Digital Interactions. It also outlines the SoMeDi approach and experiences, focusing on machine learning and artificial intelligence techniques and methodologies, including techniques for modeling and metadata searching.
Results
Phase II of the project produced one research report.
The results achieved within phase II have been disseminated in the following events: participation at ELSE 2017 conference.
Phase III – Designing the platform for Digital Interactions
Phase III covers the following activities:
- Activity 3.1 Designing the platform architecture (correlated with T1.3)
- Activity 3.2 Designing and developing the user interface (correlated with T2.1)
- Activity 3.3 Testing the platform interfaces (correlated with T2.4)
Phase description
The purpose of Phase III was to present the functionalities, uses, and architecture of the SoMeDi platform. At the same time, was performed a research study regarding technologies available for the user experience analysis; also, was described the interface design methodology, as well as the description of several solutions and testing methodologies.
Phase IV – Digital interaction platform development and testing
Phase IV includes the following activities:
- Activity 4.1 The development of the digital interactions platform
- Activity 4.2 Digital interactions tools design and development
- Activity 4.3 Testing of the digital interaction platform
Phase description
Phase IV of the SoMeDi project ran from 16th of November 2017 – 31st of May 2018 and produced one testing report and the prototype of the digital intelligence platform.
Results
The main results achieved within Phase IV are related to the collaboration between the Romanian consortium partners, BEIA and SIVECO, in order to develop the digital interaction platform.
Also, in this phase, was advanced the prototype for the DII software.
BEIA has published a paper in “Studies in Informatics and Control” Journal.
Phase V – Development of the integrated platform for digital interactions
Phase description
Phase V started on the 16th of November 2017 and ended on the 14th of November 2018 and includes the following activities:
- Activity 5.1 Developing the platform and the visualization tools
- Activity 5.2 Developing and testing the reporting tools
Results
Phase V of the project issued one testing report and the prototype of the GUI.
Phase V of the SoMeDi project carried out the following research and development activities:
o Development and deployment of the SoMeDi platform highlighting the servers and software applications necessary for its operation.
o Advanced a comparative analysis of sentiment analysis tools and proposed a methodology in order to identify the best method, respectively, Google Cloud Natural Language.
o Proposed the architecture of a web service that integrates the Google Cloud Natural Language tool for sentiment analysis, as well as the methodology for integration into the platform.
o Development of the visualization/reporting tools in the form of “Proof of Concept,” thus obtaining the validation of the sentiment analysis tool, as well as the analysis of the data flow generated by the test users registered in the SoMeDi platform.
o Presentation of dissemination support activities carried out during the current period.
The results achieved within phase V have been disseminated in the following events: participation in ECAI 2018 Conference, ICVL 2018 Conference, CSE2018 Conference, and SGEM 2018 Social Sciences and Arts Conference.
BEIA was present in the first edition of the Transylvanian Machine Learning Summer School (TMLSS).
Phase VI – Validation of the final prototype
Phase description
This phase started on the 16th of November 2018 and ended on September 1st, 2019 and includes the following activities:
- Activity 6.1 Definition of the validation cases for stakeholders
- Activity 6.2 Testing and final validation of the platform for digital interaction
- Activity 6.3 Demonstration of the solution functionality and utility
- Activity 6.4 Identification and protection of intellectual property rights
- Activity 6.5 Dissemination
Results
Phase VI of the project issued the platform users’ manual, together with the platform final testing report. In this stage was also released a report detailing the activities carried for securing the intellectual properties over the platform services.
The platform is available at http://somedi.agile.ro/
Phase VII – Reporting of economic effects (02.09.2019-01.09.2020)
Phase VIII – Reporting of economic effects (02.09.2020-01.09.2021)
Phase IX – Reporting of economic effects (02.09.2021-01.09.2022)
Social Media
Check the Twitter account here.
Papers resulted from the project:
- G. Suciu, M. Anwar, R. Conu, “Social Media and Digital Interactions using Cloud Services for orienting Young People in their Careers”, In The International Scientific Conference eLearning and Software for Education (ELSE), “Carol I” National Defence University, Vol. 2, pp. 419-427, 2017
- G. Suciu, A. Pasat, T. Usurelu, “Social Media Cloud Contact Center using ChatBots”, Fabulous 2017 – presented
- George SUCIU, Corentin BOSCHER, Laura PRIOUX, Adrian PASAT, Ciprian DOBRE, Insights into Collaborative Platforms for Social Media: Use Cases, Studies in Informatics and Control, ISSN 1220-1766, vol. 26(4), pp. 435-440, 2017 – published
- G. Suciu, A. Pasat, C. Nadrag, C. Balaceanu, A. Drosu. , “Design of an internship recruitment platform employing NLP based technologies”, ECAI 2018 INTERNATIONAL CONFERENCE 10th Edition. Electronics, Computers and Artificial Intelligence 28 June-30 June 2018, Iasi, ROMANIA. Conference Record Number:43897X – presented
- G. Suciu, C. Balaceanu, A. Pasat, C. Nadrag, “Geolocation and Social Media for Enhanced Recruitment Campaigns”, SGEM 2018; paper presented and published in the conference proceedings;
- G. Suciu, A. Pasat, M. C. Ditu, I. Rogojanu, “Smart Shopping Technologies for Indoor Markets”, CSE 2018; paper presented and accepted for publishing;
- A. Pasat, I. Rogojanu, G. Suciu, “SoMeDi: Successful Internship Programmes Matching Job Offers with Candidates Skills”, ICVL 2018, paper presented and published in the conference proceedings;
- G. Suciu, G. Bucur, R. Conu, “The importance of Social Media in Smart Cities”, URBAN INCERC 2019, paper presented and published in the conference proceedings;
- Vasilescu, C., Suciu, G., & Pasat, A. (2019). A New Method to Help the Human Resources Staff to Find the Right Candidates, Based on Deep Learning. In The International Scientific Conference eLearning and Software for Education (Vol. 3, pp. 240-246). ” Carol I” National Defence University. paper presented and published in the conference proceedings;
- G. Suciu, A. Pasat, C. Vasilescu, “Novel Artificial Intelligence Technologies for Enhanced Recruitment and Training Campaigns using Social Media”, In The International Scientific Conference eLearning and Software for Education 2019,paper presented and published in the conference proceedings;
