Description du projet
Un cadre pour calculer les réponses à des questions vastes et complexes
Aujourd’hui, l’information est plus facilement accessible dans le monde entier. Cependant, étant transmise par des systèmes automatisés, son accès est limité par la capacité de ces systèmes à comprendre la langue. Les systèmes actuels sont incapables de répondre à des questions complexes dans de vastes domaines. En effet, ils ne peuvent pas décomposer les questions en plusieurs parties et trouver les informations pertinentes dans différentes sources. Le projet DELPHI, financé par l’UE, vise à montrer que les ordinateurs peuvent répondre à des questions diverses et complexes qui nécessitent un raisonnement sur plusieurs modalités. À cette fin, il propose un cadre dans lequel les questions complexes sont symboliquement décomposées en sous-questions. Un réseau neuronal répond à chaque question, et la réponse finale est calculée à partir de toutes les informations recueillies. Ce travail a le potentiel de changer l’interaction homme-machine de l’avenir.
Objectif
The explosion of information around us has democratized knowledge and transformed its availability for
people around the world. Still, since information is mediated through automated systems, access is bounded
by their ability to understand language.
Consider an economist asking “What fraction of the top-5 growing countries last year raised their co2 emission?”.
While the required information is available, answering such complex questions automatically is
not possible. Current question answering systems can answer simple questions in broad domains, or complex
questions in narrow domains. However, broad and complex questions are beyond the reach of state-of-the-art.
This is because systems are unable to decompose questions into their parts, and find the relevant information
in multiple sources. Further, as answering such questions is hard for people, collecting large datasets to train
such models is prohibitive.
In this proposal I ask: Can computers answer broad and complex questions that require reasoning over
multiple modalities? I argue that by synthesizing the advantages of symbolic and distributed representations
the answer will be “yes”. My thesis is that symbolic representations are suitable for meaning composition, as
they provide interpretability, coverage, and modularity. Complementarily, distributed representations (learned
by neural nets) excel at capturing the fuzziness of language. I propose a framework where complex questions
are symbolically decomposed into sub-questions, each is answered with a neural network, and the final answer
is computed from all gathered information.
This research tackles foundational questions in language understanding. What is the right representation
for reasoning in language? Can models learn to perform complex actions in the face of paucity of data?
Moreover, my research, if successful, will transform how we interact with machines, and define a role for
them as research assistants in science, education, and our daily life.
Champ scientifique
Programme(s)
Thème(s)
Régime de financement
ERC-STG - Starting GrantInstitution d’accueil
69978 Tel Aviv
Israël