In the context of my PhD thesis, I worked on formal representation of NL quantifiers. In particular, I devised a new logical framework to properly represent Scopeless readings, such as cumulative and collective readings. I authored six journal publications and several conference/workshop ones on the topic and I am the single author of four of these journal publications. My last publication on the topic was
(Robaldo, Szymanik, Meijering, 2014), which are coauthored with
Jakub Szymanik and
After that publication, I stopped working on NL Quantifiers. On the other hand, Jakub was later awarded by an
ERC Starting Grant
on related topics (developing cognitive semantics of generalized quantifiers). Congrats Jakub!
Penn Discourse Treebank (PDTB)
The PDTB is a corpus developed at the University of Pennsylvania (UPenn). The PDTB is, to date, the largest annotation effort at the discourse level, providing annotations of the argument structure, attribution and semantics of discourse connectives. After my PhD thesis, I visited University of Pennsylvania for five months (and later again in 2009 for two months) where I started working with the PDTB research group. I contributed to the writing of the PDTB 2.0 annotation manual and the sense annotation in the release 2.0 of the corpus. During that period, I also started working with reification-based semantics, specifically the approach of Jerry R. Hobbs,
and I used it to model concessive relations found in the PDTB.
In 2013, I defined, together with Luigi Di Caro, an XML formalism called
OpinionMining-ML for tagging users' opinions on products and services, and I built a corpus of 1000 comments about restaurants taken from www.2spaghi.it, one of the biggest web2.0 sites about Italian restaurants and pizzerias. Afterwards, I won the
Working Capital Accelerator 2014, a Telecom Italia grant to support new startups and innovative research projects, with the project
SentiTagger, aiming at automatically tagging comments in OpinionMining-ML. The selection was highly competitive: only 40 projects out of about 1,300 submitted ones were selected. Each selected project was granted 25,000 euros from Telecom Italia.
In 2013, I worked on Gamification-based approaches to corpora building, pionereed by
Massimo Poesio. I was specifically involved in the
Phrase Detective game-with-a-purpose, aiming at creating anaphorically annotated resources through Web cooperation. I was an expert annotator of the game and I developed a converter from Italian texts to the input format of the game via dependency parsing, in order to allow annotations in Italian. Massimo Poesio was later awarded with an
ERC Advanced grant, on the project ''DALI - Disagreements and Language Interpretation'', which proposes more advanced games, drawn from Phrase Detective, to collect massive amounts of data about anaphora from people playing them. Congrats Massimo!