@incollection{fdi:010082405,
  title = {{B}alancing familiarity and curiosity in data exploration with deep reinforcement learning},
  author = {{P}ersonnaz, {A}. and {A}mer-{Y}ahia, {S}. and {B}erti-{E}quille, {L}aure and {F}abricius, {M}. and {S}ubramanian, {S}.},
  editor = {},
  language = {{ENG}},
  abstract = {{T}he ability to find a set of records in {E}xploratory {D}ata {A}nalysis ({EDA}) hinges on the scattering of objects in the data set and the onusers' knowledge of data and their ability to express their needs. {T}his yields a wide range of {EDA} scenarios and solutions that differin the guidance they provide to users. {I}n this paper, we investigate the interplay between modeling curiosity and familiarity in {D}eep {R}einforcement {L}earning ({DRL}) and expressive data exploration operators. {W}e formalize curiosity as intrinsic reward and familiarity as extrinsic reward. {W}e examine the behavior of several policies learned for different weights for those rewards. {O}ur experiments on {SDSS}, a very large sky survey data set provide several insights and justify the need for a deeper examination of combining {DRL} anddata exploration operators that go beyond drill-downs and roll-ups.},
  keywords = {{MONDE}},
  booktitle = {{F}ourth workshop in exploiting {AI} techniques for data management (ai{DM}'21)},
  numero = {},
  pages = {16--23},
  address = {{N}ew {Y}ork},
  publisher = {{ACM}},
  series = {},
  year = {2021},
  DOI = {10.1145/3464509.3464884},
  ISBN = {978-1-4503-85350},
  URL = {https://www.documentation.ird.fr/hor/fdi:010082405},
}