This deliverable reports the results achived on interaction adaptation and personalization during the first year of the project. This includes the basic version of the adaptation framework with handcrafted rules only. It also includes technical design report and first prototype implementation code for the Citizen data vault.
The project integrates a very sophisticated text adaptation engine based on a powerful set of NLP rules all of them briefly described in the report. It is very valuable that progress for Italian and Spanish languages was achieved. The workflow engine adaptation description is less developed but there is public code and demos code which the reader can use to get more details.
The deliverable is well-written and well-organized.
In particular, I appreciated the part concerning the text adaptation engine. This part is very interesting and presents innovative solutions to simplify complex sentences. To further improve the results of the proposed approach, authors could be interested in using common-sense reasoning (see Liu, Hugo, and Push Singh. "ConceptNet—a practical commonsense reasoning tool-kit." BT technology journal 22.4 (2004): 211-226.)
Also the rest of the document is good. I appreciated the reference to the General Data Protection Regulation (GDPR) in the “Citizen Data Vault” section.
The realized work is clearly described et is of interest. There are certainty improvements possible: for example, the text in figures 4 and 6 (and others) ais not readable. There are also technical questions that one could ask: can on e transfer the results obtained using the corpus Newela to the text of public administration (where the distribution of the words is rather unusual)? You present the lexical simplifiers for different languages separately. Is there no unified technique (that is adjusted based on the language)?
Overall the services (in a given language) are supposed to adapt to a group of users. How to describe/identify them
The main objectives of this WP seem quite well established and described. The result will be very important for a closer relation between citizens and Public Administration. The use of the language is a main issue, as citizens don't understand most of the acts pursued by the PA.
