Eric Gaussier

Clément Yvernes, Emilie Devijver, Marianne Clausel et al.

The do-calculus defines a general system of inference for interventional queries, allowing causal quantities to be transformed through successive applications of its rules. This process induces a rich space of equivalent interventional expressions, but combining and ordering these rules remains challenging. In this work, we introduce derivation graphs, which represent how do-calculus rules are applied and combined, and characterize the full space of observational and interventional probabilities which are equivalent under the do-calculus. The structure of these graphs yields a simple procedure that uses at most four applications of do-calculus rules. Finally, we show how applying identification algorithms to equivalent causal queries produces multiple valid estimands for the same causal quantity, eventually yielding more efficient estimators.

Minghan Li, Eric Gaussier, Juntao Li et al.

Decoder-only LLM rerankers struggle with long documents: inference is costly and relevance signals can be diluted by irrelevant context. Motivated by a diagnostic attention analysis suggesting that appended irrelevant context can weaken query-focused interactions, we propose EviRerank, an evidence-based long-document reranking framework for decoder-only LLMs. EviRerank first scores document blocks with a lightweight selector, such as BM25, a bi-encoder, or a cross-encoder. It then constructs a compact reranking context under a hard token cap by dynamically budgeting evidence blocks with Adaptive Evidence Budgeting (AEB) and adding a compact global cue via Summary Augmentation (SA). Finally, the compact evidence context is reranked with a decoder-only LLM. Across TREC DL'19, DL'22, DL'23, and MLDR-zh, EviRerank consistently outperforms full-document LLM reranking and strong block-selection baselines while reducing input length. RankZephyr-7B validation further confirms transfer to listwise reranking. On TREC DL'19, EviRerank reaches up to 0.744 nDCG@10 and 0.307 MAP, improving over RankLLaMA while using a compact evidence context.

Daria Bystrova, Charles K. Assaad, Julyan Arbel et al.

Constraint-based methods and noise-based methods are two distinct families of methods proposed for uncovering causal graphs from observational data. However, both operate under strong assumptions that may be challenging to validate or could be violated in real-world scenarios. In response to these challenges, there is a growing interest in hybrid methods that amalgamate principles from both methods, showing robustness to assumption violations. This paper introduces a novel comprehensive framework for hybridizing constraint-based and noise-based methods designed to uncover causal graphs from observational time series. The framework is structured into two classes. The first class employs a noise-based strategy to identify a super graph, containing the true graph, followed by a constraint-based strategy to eliminate unnecessary edges. In the second class, a constraint-based strategy is applied to identify a skeleton, which is then oriented using a noise-based strategy. The paper provides theoretical guarantees for each class under the condition that all assumptions are satisfied, and it outlines some properties when assumptions are violated. To validate the efficacy of the framework, two algorithms from each class are experimentally tested on simulated data, realistic ecological data, and real datasets sourced from diverse applications. Notably, two novel datasets related to Information Technology monitoring are introduced within the set of considered real datasets. The experimental results underscore the robustness and effectiveness of the hybrid approaches across a broad spectrum of datasets.

Charles K. Assaad, Emilie Devijver, Eric Gaussier et al.

We study the problem of identifiability of the total effect of an intervention from observational time series in the situation, common in practice, where one only has access to abstractions of the true causal graph. We consider here two abstractions: the extended summary causal graph, which conflates all lagged causal relations but distinguishes between lagged and instantaneous relations, and the summary causal graph which does not give any indication about the lag between causal relations. We show that the total effect is always identifiable in extended summary causal graphs and provide sufficient conditions for identifiability in summary causal graphs. We furthermore provide adjustment sets allowing to estimate the total effect whenever it is identifiable.

Ali Aït-Bachir, Charles K. Assaad, Christophe de Bignicourt et al.

Information technology (IT) systems are vital for modern businesses, handling data storage, communication, and process automation. Monitoring these systems is crucial for their proper functioning and efficiency, as it allows collecting extensive observational time series data for analysis. The interest in causal discovery is growing in IT monitoring systems as knowing causal relations between different components of the IT system helps in reducing downtime, enhancing system performance and identifying root causes of anomalies and incidents. It also allows proactive prediction of future issues through historical data analysis. Despite its potential benefits, applying causal discovery algorithms on IT monitoring data poses challenges, due to the complexity of the data. For instance, IT monitoring data often contains misaligned time series, sleeping time series, timestamp errors and missing values. This paper presents case studies on applying causal discovery algorithms to different IT monitoring datasets, highlighting benefits and ongoing challenges.

Charles K. Assaad, Emilie Devijver, Eric Gaussier

This study addresses the problem of learning an extended summary causal graph on time series. The algorithms we propose fit within the well-known constraint-based framework for causal discovery and make use of information-theoretic measures to determine (in)dependencies between time series. We first introduce generalizations of the causation entropy measure to any lagged or instantaneous relations, prior to using this measure to construct extended summary causal graphs by adapting two well-known algorithms, namely PC and FCI. The behavior of our methods is illustrated through several experiments run on simulated and real datasets.

Éloi Zablocki, Valentin Gerard, Amaia Cardiel et al.

Understanding deep models is crucial for deploying them in safety-critical applications. We introduce GIFT, a framework for deriving post-hoc, global, interpretable, and faithful textual explanations for vision classifiers. GIFT starts from local faithful visual counterfactual explanations and employs (vision) language models to translate those into global textual explanations. Crucially, GIFT provides a verification stage measuring the causal effect of the proposed explanations on the classifier decision. Through experiments across diverse datasets, including CLEVR, CelebA, and BDD, we demonstrate that GIFT effectively reveals meaningful insights, uncovering tasks, concepts, and biases used by deep vision classifiers. The framework is released at https://github.com/valeoai/GIFT.

Clément Yvernes, Emilie Devijver, Eric Gaussier

The identifiability problem for interventions aims at assessing whether the total causal effect can be written with a do-free formula, and thus be estimated from observational data only. We study this problem, considering multiple interventions, in the context of time series when only an abstraction of the true causal graph, in the form of a summary causal graph, is available. We propose in particular both necessary and sufficient conditions for the adjustment criterion, which we show is complete in this setting, and provide a pseudo-linear algorithm to decide whether the query is identifiable or not.

Lei Zan, Charles K. Assaad, Emilie Devijver et al.

This paper introduces a new structural causal model tailored for representing threshold-based IT systems and presents a new algorithm designed to rapidly detect root causes of anomalies in such systems. When root causes are not causally related, the method is proven to be correct; while an extension is proposed based on the intervention of an agent to relax this assumption. Our algorithm and its agent-based extension leverage causal discovery from offline data and engage in subgraph traversal when encountering new anomalies in online data. Our extensive experiments demonstrate the superior performance of our methods, even when applied to data generated from alternative structural causal models or real IT monitoring data.

Morad Laglil, Emilie Devijver, Eric Gaussier et al.

Inspired by recent advances in large language models, foundation models have been developed for zero-shot time series forecasting, enabling prediction on datasets unseen during pretraining. These large-scale models, trained on vast collections of time series, learn generalizable representations for both point and probabilistic forecasting, reducing the need for task-specific architectures and manual tuning. In this work, we review the main architectures, pretraining strategies, and optimization methods used in such models, and study the effect of fine-tuning after pretraining to enhance their performance on specific datasets. Our empirical results show that fine-tuning generally improves zero-shot forecasting capabilities, especially for long-term horizons.

Clément Yvernes, Emilie Devijver, Marianne Clausel et al.

Identifying the effect of a treatment from observational data typically requires assuming a fully specified causal diagram. However, such diagrams are rarely known in practice, especially in complex or high-dimensional settings. To overcome this limitation, recent works have explored the use of causal abstractions-simplified representations that retain partial causal information. In this paper, we consider causal abstractions formalized as collections of causal diagrams, and focus on the identifiability of causal queries within such collections. We introduce and formalize several identifiability criteria under this setting. Our main contribution is to organize these criteria into a structured hierarchy, highlighting their relationships. This hierarchical view enables a clearer understanding of what can be identified under varying levels of causal knowledge. We illustrate our framework through examples from the literature and provide tools to reason about identifiability when full causal knowledge is unavailable.

Clément Yvernes, Charles K. Assaad, Emilie Devijver et al.

The identifiability problem for interventions aims at assessing whether the total effect of some given interventions can be written with a do-free formula, and thus be computed from observational data only. We study this problem, considering multiple interventions and multiple effects, in the context of time series when only abstractions of the true causal graph in the form of summary causal graphs are available. We focus in this study on identifiability by a common backdoor set, and establish, for time series with and without consistency throughout time, conditions under which such a set exists. We also provide algorithms of limited complexity to decide whether the problem is identifiable or not.

Minghan Li, Diana Nicoleta Popa, Johan Chagnon et al.

On a wide range of natural language processing and information retrieval tasks, transformer-based models, particularly pre-trained language models like BERT, have demonstrated tremendous effectiveness. Due to the quadratic complexity of the self-attention mechanism, however, such models have difficulties processing long documents. Recent works dealing with this issue include truncating long documents, in which case one loses potential relevant information, segmenting them into several passages, which may lead to miss some information and high computational complexity when the number of passages is large, or modifying the self-attention mechanism to make it sparser as in sparse-attention models, at the risk again of missing some information. We follow here a slightly different approach in which one first selects key blocks of a long document by local query-block pre-ranking, and then few blocks are aggregated to form a short document that can be processed by a model such as BERT. Experiments conducted on standard Information Retrieval datasets demonstrate the effectiveness of the proposed approach.

Charles K. Assaad, Emilie Devijver, Eric Gaussier

This study addresses the problem of learning a summary causal graph on time series with potentially different sampling rates. To do so, we first propose a new causal temporal mutual information measure for time series. We then show how this measure relates to an entropy reduction principle that can be seen as a special case of the probability raising principle. We finally combine these two ingredients in PC-like and FCI-like algorithms to construct the summary causal graph. There algorithm are evaluated on several datasets, which shows both their efficacy and efficiency.

Thibaut Thonet, Yagmur Gizem Cinar, Eric Gaussier et al.

Information retrieval (IR) systems traditionally aim to maximize metrics built on rankings, such as precision or NDCG. However, the non-differentiability of the ranking operation prevents direct optimization of such metrics in state-of-the-art neural IR models, which rely entirely on the ability to compute meaningful gradients. To address this shortcoming, we propose SmoothI, a smooth approximation of rank indicators that serves as a basic building block to devise differentiable approximations of IR metrics. We further provide theoretical guarantees on SmoothI and derived approximations, showing in particular that the approximation errors decrease exponentially with an inverse temperature-like hyperparameter that controls the quality of the approximations. Extensive experiments conducted on four standard learning-to-rank datasets validate the efficacy of the listwise losses based on SmoothI, in comparison to previously proposed ones. Additional experiments with a vanilla BERT ranking model on a text-based IR task also confirm the benefits of our listwise approach.

Hamid Jalalzai, Pierre Colombo, Chloé Clavel et al.

The dominant approaches to text representation in natural language rely on learning embeddings on massive corpora which have convenient properties such as compositionality and distance preservation. In this paper, we develop a novel method to learn a heavy-tailed embedding with desirable regularity properties regarding the distributional tails, which allows to analyze the points far away from the distribution bulk using the framework of multivariate extreme value theory. In particular, a classifier dedicated to the tails of the proposed embedding is obtained which performance outperforms the baseline. This classifier exhibits a scale invariance property which we leverage by introducing a novel text generation method for label preserving dataset augmentation. Numerical experiments on synthetic and real text data demonstrate the relevance of the proposed framework and confirm that this method generates meaningful sentences with controllable attribute, e.g. positive or negative sentiment.

Xuhui Fan, Eric Gaussier

Metric learning has been successful in learning new metrics adapted to numerical datasets. However, its development on categorical data still needs further exploration. In this paper, we propose a method, called CPML for \emph{categorical projected metric learning}, that tries to efficiently~(i.e. less computational time and better prediction accuracy) address the problem of metric learning in categorical data. We make use of the Value Distance Metric to represent our data and propose new distances based on this representation. We then show how to efficiently learn new metrics. We also generalize several previous regularizers through the Schatten $p$-norm and provides a generalization bound for it that complements the standard generalization bound for metric learning. Experimental results show that our method provides

Hamid Mirisaee, Eric Gaussier, Cedric Lagnier et al.

We propose in this paper a new, hybrid document embedding approach in order to address the problem of document similarities with respect to the technical content. To do so, we employ a state-of-the-art graph techniques to first extract the keyphrases (composite keywords) of documents and, then, use them to score the sentences. Using the ranked sentences, we propose two approaches to embed documents and show their performances with respect to two baselines. With domain expert annotations, we illustrate that the proposed methods can find more relevant documents and outperform the baselines up to 27% in terms of NDCG.

Shubham Agarwal, Marc Dymetman, Eric Gaussier

This paper describes our submission to the E2E NLG Challenge. Recently, neural seq2seq approaches have become mainstream in NLG, often resorting to pre- (respectively post-) processing delexicalization (relexicalization) steps at the word-level to handle rare words. By contrast, we train a simple character level seq2seq model, which requires no pre/post-processing (delexicalization, tokenization or even lowercasing), with surprisingly good results. For further improvement, we explore two re-ranking approaches for scoring candidates. We also introduce a synthetic dataset creation procedure, which opens up a new way of creating artificial datasets for Natural Language Generation.

Myriam Tami, Marianne Clausel, Emilie Devijver et al.

Tree-based ensemble methods, as Random Forests and Gradient Boosted Trees, have been successfully used for regression in many applications and research studies. Furthermore, these methods have been extended in order to deal with uncertainty in the output variable, using for example a quantile loss in Random Forests (Meinshausen, 2006). To the best of our knowledge, no extension has been provided yet for dealing with uncertainties in the input variables, even though such uncertainties are common in practical situations. We propose here such an extension by showing how standard regression trees optimizing a quadratic loss can be adapted and learned while taking into account the uncertainties in the inputs. By doing so, one no longer assumes that an observation lies into a single region of the regression tree, but rather that it belongs to each region with a certain probability. Experiments conducted on several data sets illustrate the good behavior of the proposed extension.

Maziar Moradi Fard, Thibaut Thonet, Eric Gaussier

We study in this paper the problem of jointly clustering and learning representations. As several previous studies have shown, learning representations that are both faithful to the data to be clustered and adapted to the clustering algorithm can lead to better clustering performance, all the more so that the two tasks are performed jointly. We propose here such an approach for $k$-Means clustering based on a continuous reparametrization of the objective function that leads to a truly joint solution. The behavior of our approach is illustrated on various datasets showing its efficacy in learning representations for objects while clustering them.

Yagmur G. Cinar, Hamid Mirisaee, Parantapa Goswami et al.

We propose here an extended attention model for sequence-to-sequence recurrent neural networks (RNNs) designed to capture (pseudo-)periods in time series. This extended attention model can be deployed on top of any RNN and is shown to yield state-of-the-art performance for time series forecasting on several univariate and multivariate time series.

Ioannis Partalas, Aris Kosmopoulos, Nicolas Baskiotis et al.

LSHTC is a series of challenges which aims to assess the performance of classification systems in large-scale classification in a a large number of classes (up to hundreds of thousands). This paper describes the dataset that have been released along the LSHTC series. The paper details the construction of the datsets and the design of the tracks as well as the evaluation measures that we implemented and a quick overview of the results. All of these datasets are available online and runs may still be submitted on the online server of the challenges.

Abdelkader El Mahdaouy, Saïd EL Alaoui Ouatik, Eric Gaussier

Automatic Multi-Word Term (MWT) extraction is a very important issue to many applications, such as information retrieval, question answering, and text categorization. Although many methods have been used for MWT extraction in English and other European languages, few studies have been applied to Arabic. In this paper, we propose a novel, hybrid method which combines linguistic and statistical approaches for Arabic Multi-Word Term extraction. The main contribution of our method is to consider contextual information and both termhood and unithood for association measures at the statistical filtering step. In addition, our technique takes into account the problem of MWT variation in the linguistic filtering step. The performance of the proposed statistical measure (NLC-value) is evaluated using an Arabic environment corpus by comparing it with some existing competitors. Experimental results show that our NLC-value measure outperforms the other ones in term of precision for both bi-grams and tri-grams.

Aris Kosmopoulos, Ioannis Partalas, Eric Gaussier et al.

Hierarchical classification addresses the problem of classifying items into a hierarchy of classes. An important issue in hierarchical classification is the evaluation of different classification algorithms, which is complicated by the hierarchical relations among the classes. Several evaluation measures have been proposed for hierarchical classification using the hierarchy in different ways. This paper studies the problem of evaluation in hierarchical classification by analyzing and abstracting the key components of the existing performance measures. It also proposes two alternative generic views of hierarchical evaluation and introduces two corresponding novel measures. The proposed measures, along with the state-of-the art ones, are empirically tested on three large datasets from the domain of text classification. The empirical results illustrate the undesirable behavior of existing approaches and how the proposed methods overcome most of these methods across a range of cases.

Frank Meyer, Françoise Fessant, Fabrice Clérot et al.

In this paper, we propose an approach to analyze the performance and the added value of automatic recommender systems in an industrial context. We show that recommender systems are multifaceted and can be organized around 4 structuring functions: help users to decide, help users to compare, help users to discover, help users to explore. A global off line protocol is then proposed to evaluate recommender systems. This protocol is based on the definition of appropriate evaluation measures for each aforementioned function. The evaluation protocol is discussed from the perspective of the usefulness and trust of the recommendation. A new measure called Average Measure of Impact is introduced. This measure evaluates the impact of the personalized recommendation. We experiment with two classical methods, K-Nearest Neighbors (KNN) and Matrix Factorization (MF), using the well known dataset: Netflix. A segmentation of both users and items is proposed to finely analyze where the algorithms perform well or badly. We show that the performance is strongly dependent on the segments and that there is no clear correlation between the RMSE and the quality of the recommendation.