TERENCE is a 3-year long project designing the first adaptive learning system for poor comprehenders, hearing and deaf, and for their educators.
TERENCE in a nutshell
The learning material of TERENCE is made of books of stories and smart games for reasoning about stories, divided into difficult levels. Relaxing games accompany smart games and stimulate visuo-perceptual skills useful for interacting with the TERENCE system.
For designing useful and accessible material, the project consortium developed guidelines for writing and simplifying stories, as well as for designing games, called “smart games”, for its learners, that is, primary school children from Italy and UK.
From a technological view-point, the project developed an authoring tool for the semi-automated generation of smart games from narratives, as well as a rule-based adaptive engine for adapting its learning material to its learners.
The profiles of the TERENCE learners were created after an extensive analysis of the requirements of the learners, involving more than 500 children across Italy and UK. The profiles were revised via evaluation experiments with learners, educators and domain experts, through which incrementally improved releases of the TERENCE system were evaluated so as to attain its usability and pedagogical effectiveness. As for effectiveness, a statistical significant improvement in reading comprehension was highlighted.
The final review from the European Commission also reported an excellent progress.
- University of L’Aquila (Italy)
- University of Bozen-Bolzano (Italy)
- Moholy-Nagy Művészeti Egyetem – Hírek (Hungary)
- University of Leuven – KU Leuven (Belgium)
- Fondazione Bruno Kessler (Italy)
- University of Sussex (UK)
- SIVECO (Romania)
- Leibniz Universität Hannover – L3S Research Center (Germany)
- University of Verona (Italy)
- University of Salamanca (Spain)
- AMNIM (Slovenia)
- University of Padova (Italy)
Workpackages and Deliverables
- WP1 will take into consideration the results of other WPs, in particular, it will intertwine with WP7 (evaluation). WP1 aims at analysing the context of use and the user requirements. First, we will analyse the context of use and the TERENCE users in order to gain insights into the user needs. The analysis will be done with standardised assessments (e.g., NARA), and ad-hoc designed tests, common to English and Italian. The resulting needs will drive the specifications of the user requirements. In turn, the identified user requirements will guide the design of the TERENCE system (e.g., the design of its smart games as in WP2), and provide the basis for the evaluation WP (WP7). In this manner, WP1 will result in a unified set of functional and non-functional requirements for the scientific and technical components of TERENCE.Deliverables – D1.1, D1.2.1, D1.2.2
- WP2 is concerned with the design of the Adaptive Learning System (ALS) within TERENCE, in particular, its conceptual model, realised as ontologies, and adaptation engine. Software design is the process of problem solving and planning for a software solution. After the purpose and requirements of TERENCE are determined (WP1), WP2 produces a design of the system, more precisely, the conceptual model, the design of the architecture, the design of the GUI, and the design of the adaptive engine.Also stories and games need to be modelled, and information related to a student’s learning processes need to be stored. The following general approach will be pursued: (1) the requirements set by WP1 will allow us to model individual stories and games (part of the domain model of the conceptual model); (2) following the classification of users resulting from WP1, each learner’s competencies in reproducing relevant aspects of a story are assessed automatically, by also including the educator’s feedback (i.e., student model of the conceptual model); (3) based on these models, students are redirected to appropriate stories and games (as specified by the adaptive model of the conceptual model). The employment of an innovative automated engine: the smart games of TERENCE with the intelligent feedback will be generated by the automated reasoning module, and the reasoner’s feedback with explanation can then be employed by the progress-based guidance mechanism of the TERENCE adaptive engine.Deliverables – D2.1, D2.2, D2.3, D2.4
- The goal of this WP is to provide the underlying technology for the intuitive and advanced interfaces and interaction styles specialised for the TERENCE users. This initially requires to extend the ISO-TimeML annotation language so as to take into account the requirements sets in WP1, and the design of smart games in WP2. The extension will result in a novel temporal annotation language, that will enable WP4 to generate the smart games that use the annotated stories (e.g. who the agent of a certain event is). Then, WP3 will also use the technologies and tools currently available for the extended ISO-TimeML language, in terms of two annotation modules, for the English and Italian languages (Task 3.2). Finally, by using the annotation modules, WP3 will annotate the stories written in WP2 (Task 3.3) in terms of their events, (causal-)temporal relations between them, and participants of the events.Deliverables – D3.1, D3.2, D3.3
- The final goal of this work package is to provide the underlying technology for releasing smart games and an intelligent feedback for smart games (Task 4.1) to educators and learners, using the annotated stories. Thus this WP interacts with the annotation process of WP3, integrating information that is implicit and not automatically deduced in the annotation process of WP3 (Task 4.2). WP4 then takes as input the annotations for creating smart games that are designed in WP2 (Task 4.3). Specific care will be given to the feedback: the challenging goal is to provide automated explanatory feedback, adequate to poor comprehenders.Deliverables – D4.1, D4.2, D4.3
- The focus of TERENCE is on intuitive and innovative forms of reading and reasoning interaction. For reading and reasoning, illustrations in line with the capabilities of the target groups and the objective of improving reading comprehension are crucial. Accordingly, this WP will mostly focus on the design and the production of the large number of illustrations needed in stories, games, and all related items (like avatars, maps, characters, events, ID cards). Accordingly, the goal of this work package is twofold: (i) provide the images for the stories, games and their related items, (ii) develop the set of relaxing games designed in WP2.Deliverables – D5.1, D5.2
- Starting from the design solutions of WP2, this WP will take care of their implementation, namely: the global system architecture and the GUIs. Furthermore, this WP will also design and develop the visualisation module, and will develop the software infrastructure that enables the integration of all TERENCE modules. Finally, this WP will take care of integrating (i) the annotation, reasoning, and visualisation modules of the TERENCE system into an application with user-friendly GUIs, and (ii) the developed stories and games with their illustrationsDeliverables – D6.1, D6.2
- The evaluation of a tool is an essential step in the UCD, and is the main objective of WP7: conducting evaluations necessary for WP1 and, more in general, for the scientific and technological development of the TERENCE ALS. In general, designers need to evaluate the whole system and each of its core components in order to assess its usability, where usability is the extent to which a product can be used adequately, with satisfaction and effectiveness by specific users to achieve specific goals in a specific environment (ISO, 1989). In the TERENCE project, we will evaluate the usability and the pedagogical effectiveness of the overall ALS, and of prototypes of stories and smart games in particular. UEMs can be used as expert-based evaluation methods (e.g., heuristic evaluation and cognitive walkthrough) and user-based evaluation methods (e.g., observational evaluations, controlled evaluations, inquiries). The former methods require that experts evaluate a prototype of the system, comparing it to existing rules and guidelines. Expert-based criteria are often criticised because they depend upon the quality of the expert(s). User-based evaluations complement essentially expert-based methods, in that, user-based evaluations involve the intended end-users of the system, that use or comment on a prototype of the system. In particular, the evaluations testing the pedagogical effectiveness of TERENCE will be carefully designed in Task 7.2. Another objective of WP7 is to design evaluation guidelines for evaluators that are in charge of designing devices, environments and methods for the TERENCE learners. In particular, the evaluations for the pedagogical effectiveness of TERENCE will employ two groups of users: one group will use TERENCE; the other will not. We will conduct pre and post-tests in order to measure the actual text comprehension, and evaluate their results statistically. The stimulation plan will be refined in Task 7.2. Another objective of WP7 is to design evaluation guidelines for evaluators that are in charge of designing devices, environments and methods for the TERENCE learners.Deliverables – D7.1, D7.2, D7.3, D7.4
- The project success is based on a proper dissemination strategy. Identifying the best means of promoting and creating a “marketing network” will increase the success of the project. This WP will provide dissemination and exploitation support for TERENCE. Our dissemination work will focus on creating an overall strategy for dissemination, creating and maintaining a TERENCE website, organising and tracking all TERENCE dissemination events.Deliverables – D8.1, D8.2, D8.3, D8.4, D8.5, D8.6, D8.7
- The overall project responsibility will be shared by the management team consisting of partners represented in the Project Technical Board and in the Project Executive Board (PEB). The two functions focus on the areas that are relevant for the success of the overall project, namely scientific and technological coordination and efficient and effective administrative support and service. The WP9 objectives are to:Deliverables – No public deliverables
1. Assure the coherence of the individual WPs within the detailed project plan, the continued compliance of the WPs’ with the project objectives, and the continued relevance of the WPs’ deliverables to the project objectives.
2. Develop and maintain an efficient support and service structure for regular communications with the European Commission, reporting, and cost claiming.
3. Monitor the project integrity for quality assurance and risk management, through the definition and execution of a Critical Success Factor (CSF) and Risk management plan.
The TERENCE software
- The Learner’s GUI is where the learners started interacting with TERENCE
- Watch the video that shows how TERENCE works
TERENCE wants you 🙂
Carissimi docenti delle scuole primarie e non, come saprete dalla vostra visita al sito, TERENCE è uno strumento rivolto a quei bambini con cattiva comprensione del testo che affollano le vostre classi; come docenti sapete che avere degli strumenti innovativi che vi supportano è estremamente importante per voi e sapete che senza il vostro supporto nessuno strumento, neanche TERENCE, potrebbe funzionare. Per questo per chi fosse interessato a collaborare con noi al miglioramento di TERENCE, all’arricchimento delle storie usate per rafforzare la comprensione del testo scritto o semplicemente fosse curioso di saperne di più può contattare:
- A L’Aquila:
Pierpaolo Vittorini <pierpaolo.vittorini @ univaq.it>
Tania Di Mascio <tania.dimascio @ univaq.it>
- A Bolzano:
Rosella Gennari <gennari @ inf.unibz.it>