Marie Guegan

Scene reordering in movie script alignment

Screenplays are a very interesting source of information to describe and analyze a movie. Once aligned to the video timeline using time information extracted from the subtitles, they can be useful in many multimedia applications because they contain textual descriptions of scenes, names of the speaking characters, dialogues, etc. An alignment however is not always straightforward. The scripts freely available online often do not reflect the released version of the movie showing changes in dialogues, as well as scenes added, removed or reordered in post-production. In this paper we present two improvements to the algorithm traditionally used to align movie scripts with their videos. In particular we show that performance can be improved when taking into account the sequences that were reordered during the editing phase. We also present an automatic categorization of screenplays based on their similarity with the released movie.

Role detection in online forums based on growth models for trees

Some structural characteristics of online discussions have been successfully modeled in the recent years. When parameters of these models are properly estimated, the models are able to generate synthetic discussions that are structurally similar to the real discussions. A common aspect of these models is that they consider that all users behave according to the same model. In this paper, we combine a growth model with an Expectation–Maximization algorithm that finds different parameters for different latent groups of users. We use this method to find the different roles that coexist in the community. Moreover, we analyze whether we can predict users behaviors based on their roles. Indeed, we show that predictions are improved for some of the roles when compared with a simple growth model.

Non-parametric clustering over user features and latent behavioral functions with dual-view mixture models

We present a dual-view mixture model to cluster users based on their features and latent behavioral functions. Every component of the mixture model represents a probability density over a feature view for observed user attributes and a behavior view for latent behavioral functions that are indirectly observed through user actions or behaviors. Our task is to infer the groups of users as well as their latent behavioral functions. We also propose a non-parametric version based on a Dirichlet Process to automatically infer the number of clusters. We test the properties and performance of the model on a synthetic dataset that represents the participation of users in the threads of an online forum. Experiments show that dual-view models outperform single-view ones when one of the views lacks information.