arXiv subjects

Mariano Beguerisse-Díaz

Publications and source records attributed to Mariano Beguerisse-Díaz.

17 recordsLinked to original sources

arXiv2026

Audio-Based Understanding of Audiobook Narration Appeal

Narration is central to the audiobook listening experience, shaping how listeners engage with and understand the content. This work explores how narration qualities shape an audiobook's appeal, noting that their effects can vary by genre, title, and audience. We extract vocal and acoustic features (e.g., tone, pace, loudness) from LibriVox using pre-trained audio models and analyse their relationship with consumption data (specifically, view-rate) and their interplay with genre and title. Despite limited consumption data, we find that acoustic information alone has a robust association with appeal, even after accounting for title effects. We further validate these findings using more nuanced proprietary engagement metrics. To our knowledge, this is the first systematic computational study linking narration qualities, genre, title, and audiobook consumption, highlighting the potential of data-driven insights to improve audiobook personalisation and narrator casting.

cs.CL↗

arXiv2024

Classification of Spontaneous and Scripted Speech for Multilingual Audio

Distinguishing scripted from spontaneous speech is an essential tool for better understanding how speech styles influence speech processing research. It can also improve recommendation systems and discovery experiences for media users through better segmentation of large recorded speech catalogues. This paper addresses the challenge of building a classifier that generalises well across different formats and languages. We systematically evaluate models ranging from traditional, handcrafted acoustic and prosodic features to advanced audio transformers, utilising a large, multilingual proprietary podcast dataset for training and validation. We break down the performance of each model across 11 language groups to evaluate cross-lingual biases. Our experimental analysis extends to publicly available datasets to assess the models' generalisability to non-podcast domains. Our results indicate that transformer-based models consistently outperform traditional feature-based techniques, achieving state-of-the-art performance in distinguishing between scripted and spontaneous speech across various languages.

cs.CL↗

arXiv2024

Link Me Baby One More Time: Social Music Discovery on Spotify

We explore the social and contextual factors that influence the outcome of person-to-person music recommendations and discovery. Specifically, we use data from Spotify to investigate how a link sent from one user to another results in the receiver engaging with the music of the shared artist. We consider several factors that may influence this process, such as the strength of the sender-receiver relationship, the user's role in the Spotify social network, their music social cohesion, and how similar the new artist is to the receiver's taste. We find that the receiver of a link is more likely to engage with a new artist when (1) they have similar music taste to the sender and the shared track is a good fit for their taste, (2) they have a stronger and more intimate tie with the sender, and (3) the shared artist is popular amongst the receiver's connections. Finally, we use these findings to build a Random Forest classifier to predict whether a shared music track will result in the receiver's engagement with the shared artist. This model elucidates which type of social and contextual features are most predictive, although peak performance is achieved when a diverse set of features are included. These findings provide new insights into the multifaceted mechanisms underpinning the interplay between music discovery and social processes.

cs.SI↗

arXiv2023

Topological fingerprints for audio identification

We present a topological audio fingerprinting approach for robustly identifying duplicate audio tracks. Our method applies persistent homology on local spectral decompositions of audio signals, using filtered cubical complexes computed from mel-spectrograms. By encoding the audio content in terms of local Betti curves, our topological audio fingerprints enable accurate detection of time-aligned audio matchings. Experimental results demonstrate the accuracy of our algorithm in the detection of tracks with the same audio content, even when subjected to various obfuscations. Our approach outperforms existing methods in scenarios involving topological distortions, such as time stretching and pitch shifting.

cs.SD↗

arXiv2021

Opinion dynamics on tie-decay networks

In social networks, interaction patterns typically change over time. We study opinion dynamics on tie-decay networks in which tie strength increases instantaneously when there is an interaction and decays exponentially between interactions. Specifically, we formulate continuous-time Laplacian dynamics and a discrete-time DeGroot model of opinion dynamics on these tie-decay networks, and we carry out numerical computations for the continuous-time Laplacian dynamics. We examine the speed of convergence by studying the spectral gaps of combinatorial Laplacian matrices of tie-decay networks. First, we compare the spectral gaps of the Laplacian matrices of tie-decay networks that we construct from empirical data with the spectral gaps for corresponding randomized and aggregate networks. We find that the spectral gaps for the empirical networks tend to be smaller than those for the randomized and aggregate networks. Second, we study the spectral gap as a function of the tie-decay rate and time. Intuitively, we expect small tie-decay rates to lead to fast convergence because the influence of each interaction between two nodes lasts longer for smaller decay rates. Moreover, as time progresses and more interactions occur, we expect eventual convergence. However, we demonstrate that the spectral gap need not decrease monotonically with respect to the decay rate or increase monotonically with respect to time. Our results highlight the importance of the interplay between the times that edges strengthen and decay in temporal networks.

cs.SI↗

arXiv2021

Thematic recommendations on knowledge graphs using multilayer networks

We present a framework to generate and evaluate thematic recommendations based on multilayer network representations of knowledge graphs (KGs). In this representation, each layer encodes a different type of relationship in the KG, and directed interlayer couplings connect the same entity in different roles. The relative importance of different types of connections is captured by an intuitive salience matrix that can be estimated from data, tuned to incorporate domain knowledge, address different use cases, or respect business logic. We apply an adaptation of the personalised PageRank algorithm to multilayer models of KGs to generate item-item recommendations. These recommendations reflect the knowledge we hold about the content and are suitable for thematic and/or cold-start recommendation settings. Evaluating thematic recommendations from user data presents unique challenges that we address by developing a method to evaluate recommendations relying on user-item ratings, yet respecting their thematic nature. We also show that the salience matrix can be estimated from user data. We demonstrate the utility of our methods by significantly improving consumption metrics in an AB test where collaborative filtering delivered subpar performance. We also apply our approach to movie recommendation using publicly-available data to ensure the reproducibility of our results. We demonstrate that our approach outperforms existing thematic recommendation methods and is even competitive with collaborative filtering approaches.

cs.IR↗

arXiv2021

Tie-decay networks in continuous time and eigenvector-based centralities

Network theory is a useful framework for studying interconnected systems of interacting entities. Many networked systems evolve continuously in time, but most existing methods for the analysis of time-dependent networks rely on discrete or discretized time. In this paper, we propose an approach for studying networks that evolve in continuous time by distinguishing between \emph{interactions}, which we model as discrete contacts, and \emph{ties}, which encode the strengths of relationships as functions of time. To illustrate our tie-decay network formalism, we adapt the well-known PageRank centrality score to our tie-decay framework in a mathematically tractable and computationally efficient way. We apply this framework to a synthetic example and then use it to study a network of retweets during the 2012 National Health Service controversy in the United Kingdom. Our work also provides guidance for similar generalizations of other tools from network theory to continuous-time networks with tie decay, including for applications to streaming data.

physics.soc-ph↗

arXiv2020

Automation and occupational mobility: A data-driven network model

The potential impact of automation on the labor market is a topic that has generated significant interest and concern amongst scholars, policymakers, and the broader public. A number of studies have estimated occupation-specific risk profiles by examining the automatability of associated skills and tasks. However, relatively little work has sought to take a more holistic view on the process of labor reallocation and how employment prospects are impacted as displaced workers transition into new jobs. In this paper, we develop a new data-driven model to analyze how workers move through an empirically derived occupational mobility network in response to automation scenarios which increase labor demand for some occupations and decrease it for others. At the macro level, our model reproduces a key stylized fact in the labor market known as the Beveridge curve and provides new insights for explaining the curve's counter-clockwise cyclicality. At the micro level, our model provides occupation-specific estimates of changes in short and long-term unemployment corresponding to a given automation shock. We find that the network structure plays an important role in determining unemployment levels, with occupations in particular areas of the network having very few job transition opportunities. Such insights could be fruitfully applied to help design more efficient and effective policies aimed at helping workers adapt to the changing nature of the labor market.

econ.GN↗

arXiv2019

Customer mobility and congestion in supermarkets

The analysis and characterization of human mobility using population-level mobility models is important for numerous applications, ranging from the estimation of commuter flows in cities to modeling trade flows between countries. However, almost all of these applications have focused on large spatial scales, which typically range between intra-city scales to inter-country scales. In this paper, we investigate population-level human mobility models on a much smaller spatial scale by using them to estimate customer mobility flow between supermarket zones. We use anonymized, ordered customer-basket data to infer empirical mobility flow in supermarkets, and we apply variants of the gravity and intervening-opportunities models to fit this mobility flow and estimate the flow on unseen data. We find that a doubly-constrained gravity model and an extended radiation model (which is a type of intervening-opportunities model) can successfully estimate 65--70\% of the flow inside supermarkets. Using a gravity model as a case study, we then investigate how to reduce congestion in supermarkets using mobility models. We model each supermarket zone as a queue, and we use a gravity model to identify store layouts with low congestion, which we measure either by the maximum number of visits to a zone or by the total mean queue size. We then use a simulated-annealing algorithm to find store layouts with lower congestion than a supermarket's original layout. In these optimized store layouts, we find that popular zones are often in the perimeter of a store. Our research gives insight both into how customers move in supermarkets and into how retailers can arrange stores to reduce congestion. It also provides a case study of human mobility on small spatial scales.

physics.soc-ph↗

arXiv2019

Unravelling the forces underlying urban industrial agglomeration

As early as the 1920's Marshall suggested that firms co-locate in cities to reduce the costs of moving goods, people, and ideas. These 'forces of agglomeration' have given rise, for example, to the high tech clusters of San Francisco and Boston, and the automobile cluster in Detroit. Yet, despite its importance for city planners and industrial policy-makers, until recently there has been little success in estimating the relative importance of each Marshallian channel to the location decisions of firms. Here we explore a burgeoning literature that aims to exploit the co-location patterns of industries in cities in order to disentangle the relationship between industry co-agglomeration and customer/supplier, labour and idea sharing. Building on previous approaches that focus on across- and between-industry estimates, we propose a network-based method to estimate the relative importance of each Marshallian channel at a meso scale. Specifically, we use a community detection technique to construct a hierarchical decomposition of the full set of industries into clusters based on co-agglomeration patterns, and show that these industry clusters exhibit distinct patterns in terms of their relative reliance on individual Marshallian channels.

econ.GN↗

arXiv2019

Tensor clustering with algebraic constraints gives interpretable groups of crosstalk mechanisms in breast cancer

We introduce a tensor-based clustering method to extract sparse, low-dimensional structure from high-dimensional, multi-indexed datasets. This framework is designed to enable detection of clusters of data in the presence of structural requirements which we encode as algebraic constraints in a linear program. Our clustering method is general and can be tailored to a variety of applications in science and industry. We illustrate our method on a collection of experiments measuring the response of genetically diverse breast cancer cell lines to an array of ligands. Each experiment consists of a cell line-ligand combination, and contains time-course measurements of the early-signalling kinases MAPK and AKT at two different ligand dose levels. By imposing appropriate structural constraints and respecting the multi-indexed structure of the data, the analysis of clusters can be optimized for biological interpretation and therapeutic understanding. We then perform a systematic, large-scale exploration of mechanistic models of MAPK-AKT crosstalk for each cluster. This analysis allows us to quantify the heterogeneity of breast cancer cell subtypes, and leads to hypotheses about the signalling mechanisms that mediate the response of the cell lines to ligands.

q-bio.QM↗

arXiv2018

Flux-dependent graphs for metabolic networks

Cells adapt their metabolic fluxes in response to changes in the environment. We present a framework for the systematic construction of flux-based graphs derived from organism-wide metabolic networks. Our graphs encode the directionality of metabolic fluxes via edges that represent the flow of metabolites from source to target reactions. The methodology can be applied in the absence of a specific biological context by modelling fluxes probabilistically, or can be tailored to different environmental conditions by incorporating flux distributions computed through constraint-based approaches such as Flux Balance Analysis. We illustrate our approach on the central carbon metabolism of Escherichia coli and on a metabolic model of human hepatocytes. The flux-dependent graphs under various environmental conditions and genetic perturbations exhibit systemic changes in their topological and community structure, which capture the re-routing of metabolic fluxes and the varying importance of specific reactions and pathways. By integrating constraint-based models and tools from network science, our framework allows the study of context-specific metabolic responses at a system level beyond standard pathway descriptions.

q-bio.MN↗

arXiv2017

Integrating sentiment and social structure to determine preference alignments: The Irish Marriage Referendum

We examine the relationship between social structure and sentiment through the analysis of a large collection of tweets about the Irish Marriage Referendum of 2015. We obtain the sentiment of every tweet with the hashtags #marref and #marriageref that was posted in the days leading to the referendum, and construct networks to aggregate sentiment and use it to study the interactions among users. Our results show that the sentiment of mention tweets posted by users is correlated with the sentiment of received mentions, and there are significantly more connections between users with similar sentiment scores than among users with opposite scores in the mention and follower networks. We combine the community structure of the two networks with the activity level of the users and sentiment scores to find groups of users who support voting `yes' or `no' in the referendum. There were numerous conversations between users on opposing sides of the debate in the absence of follower connections, which suggests that there were efforts by some users to establish dialogue and debate across ideological divisions. Our analysis shows that social structure can be integrated successfully with sentiment to analyse and understand the disposition of social media users. These results have potential applications in the integration of data and meta-data to study opinion dynamics, public opinion modelling, and polling.

cs.SI↗

arXiv2017

The 'who' and 'what' of #diabetes on Twitter

Social media are being increasingly used for health promotion, yet the landscape of users, messages and interactions in such fora is poorly understood. Studies of social media and diabetes have focused mostly on patients, or public agencies addressing it, but have not looked broadly at all the participants or the diversity of content they contribute. We study Twitter conversations about diabetes through the systematic analysis of 2.5 million tweets collected over 8 months and the interactions between their authors. We address three questions: (1) what themes arise in these tweets?, (2) who are the most influential users?, (3) which type of users contribute to which themes? We answer these questions using a mixed-methods approach, integrating techniques from anthropology, network science and information retrieval such as thematic coding, temporal network analysis, and community and topic detection. Diabetes-related tweets fall within broad thematic groups: health information, news, social interaction, and commercial. At the same time, humorous messages and references to popular culture appear consistently, more than any other type of tweet. We classify authors according to their temporal 'hub' and 'authority' scores. Whereas the hub landscape is diffuse and fluid over time, top authorities are highly persistent across time and comprise bloggers, advocacy groups and NGOs related to diabetes, as well as for-profit entities without specific diabetes expertise. Top authorities fall into seven interest communities as derived from their Twitter follower network. Our findings have implications for public health professionals and policy makers who seek to use social media as an engagement tool and to inform policy design.

physics.soc-ph↗

arXiv2016

Flow-based network analysis of the Caenorhabditis elegans connectome

We exploit flow propagation on the directed neuronal network of the nematode Caenorhabditis elegans to reveal dynamically relevant features of its connectome. We find flow-based groupings of neurons at different levels of granularity, which we relate to functional and anatomical constituents of its nervous system. A systematic in silico evaluation of the full set of single and double neuron ablations is used to identify deletions that induce the most severe disruptions of the multi-resolution flow structure. Such ablations are linked to functionally relevant neurons, and suggest potential candidates for further in vivo investigation. In addition, we use the directional patterns of incoming and outgoing network flows at all scales to identify flow profiles for the neurons in the connectome, without pre-imposing a priori categories. The four flow roles identified are linked to signal propagation motivated by biological input-response scenarios.

q-bio.NC↗

arXiv2014

Interest communities and flow roles in directed networks: the Twitter network of the UK riots

Directionality is a crucial ingredient in many complex networks in which information, energy or influence are transmitted. In such directed networks, analysing flows (and not only the strength of connections) is crucial to reveal important features of the network that might go undetected if the orientation of connections is ignored. We showcase here a flow-based approach for community detection in networks through the study of the network of the most influential Twitter users during the 2011 riots in England. Firstly, we use directed Markov Stability to extract descriptions of the network at different levels of coarseness in terms of interest communities, i.e., groups of nodes within which flows of information are contained and reinforced. Such interest communities reveal user groupings according to location, profession, employer, and topic. The study of flows also allows us to generate an interest distance, which affords a personalised view of the attention in the network as viewed from the vantage point of any given user. Secondly, we analyse the profiles of incoming and outgoing long-range flows with a combined approach of role-based similarity and the novel relaxed minimum spanning tree algorithm to reveal that the users in the network can be classified into five roles. These flow roles go beyond the standard leader/follower dichotomy and differ from classifications based on regular/structural equivalence. We then show that the interest communities fall into distinct informational organigrams characterised by a different mix of user roles reflecting the quality of dialogue within them. Our generic framework can be used to provide insight into how flows are generated, distributed, preserved and consumed in directed networks.

physics.soc-ph↗

arXiv2013

Finding role communities in directed networks using Role-Based Similarity, Markov Stability and the Relaxed Minimum Spanning Tree

We present a framework to cluster nodes in directed networks according to their roles by combining Role-Based Similarity (RBS) and Markov Stability, two techniques based on flows. First we compute the RBS matrix, which contains the pairwise similarities between nodes according to the scaled number of in- and out-directed paths of different lengths. The weighted RBS similarity matrix is then transformed into an undirected similarity network using the Relaxed Minimum-Spanning Tree (RMST) algorithm, which uses the geometric structure of the RBS matrix to unblur the network, such that edges between nodes with high, direct RBS are preserved. Finally, we partition the RMST similarity network into role-communities of nodes at all scales using Markov Stability to find a robust set of roles in the network. We showcase our framework through a biological and a man-made network.

cs.SI↗