arXiv subjects

Tiago Cunha

Publications and source records attributed to Tiago Cunha.

9 recordsLinked to original sources

arXiv2026

LLM-Based Re-Ranking for Real Estate Search

QuintoAndar Group operates the leading housing marketplace in Latin America for both rentals and sales. The platform replaces traditionally paper-heavy workflows with a fully digital experience, making housing transactions faster and more accessible to tenants, buyers, and landlords in the region. Finding the ideal home in such a vast catalog is inherently difficult. At the same time, the widespread adoption of conversational assistants is reshaping user expectations: people increasingly want to express their needs through open, multi-turn dialog rather than rigid filter menus and faceted search. This shift is particularly pronounced in housing, where intent is multi-dimensional, context-dependent, and rarely reducible to a small set of structured constraints. To meet these expectations, we propose a Large Language Model (LLM) based re-ranker that augments a conversational recommendation system by reordering retrieved candidates according to the nuanced, context-rich intent expressed across the user's conversation. We additionally construct a large-scale offline evaluation dataset for conversational real-estate search, containing 960,000 query-item pairs constructed from both synthetic and production queries and annotated using an LLM-as-a-Judge framework with human validation. We validate our approach both offline, on this proprietary dataset, and online, through a production A/B test. Both evaluations show consistent improvements in ranking quality, including a statistically significant increase in production of +5.3% in click-through rate and +4.8% in scheduled visits, demonstrating the value of integrating conversational context into housing recommendations.

cs.IR↗

arXiv2024

A Hybrid Meta-Learning and Multi-Armed Bandit Approach for Context-Specific Multi-Objective Recommendation Optimization

Recommender systems in online marketplaces face the challenge of balancing multiple objectives to satisfy various stakeholders, including customers, providers, and the platform itself. This paper introduces Juggler-MAB, a hybrid approach that combines meta-learning with Multi-Armed Bandits (MAB) to address the limitations of existing multi-stakeholder recommendation systems. Our method extends the Juggler framework, which uses meta-learning to predict optimal weights for utility and compensation adjustments, by incorporating a MAB component for real-time, context-specific refinements. We present a two-stage approach where Juggler provides initial weight predictions, followed by MAB-based adjustments that adapt to rapid changes in user behavior and market conditions. Our system leverages contextual features such as device type and brand to make fine-grained weight adjustments based on specific segments. To evaluate our approach, we developed a simulation framework using a dataset of 0.6 million searches from Expedia's lodging booking platform. Results show that Juggler-MAB outperforms the original Juggler model across all metrics, with NDCG improvements of 2.9%, a 13.7% reduction in regret, and a 9.8% improvement in best arm selection rate.

cs.LG↗

arXiv2024

Early Career Wins and Tournament Prestige Characterize Tennis Players' Trajectories

Success in sports is a complex phenomenon that has only garnered limited research attention. In particular, we lack a deep scientific understanding of success in sports like tennis and the factors that contribute to it. Here, we study the unfolding of tennis players' careers to understand the role of early career stages and the impact of specific tournaments on players' trajectories. We employ a comprehensive approach combining network science and analysis of ATP tournament data and introduce a novel method to quantify tournament prestige based on the eigenvector centrality of the co-attendance network of tournaments. Focusing on the interplay between participation in central tournaments and players' performance, we find that the level of the tournament where players achieve their first win is associated with becoming a top player. This work sheds light on the critical role of the initial stages in the progression of players' careers, offering valuable insights into the dynamics of success in tennis.

physics.soc-ph↗

arXiv2022

Automated Detection of Missing Links in Bicycle Networks

Cycling is an effective solution for making urban transport more sustainable. However, bicycle networks are typically developed in a slow, piecewise process that leaves open a large number of gaps, even in well developed cycling cities like Copenhagen. Here, we develop the IPDC procedure (Identify, Prioritize, Decluster, Classify) for finding the most important missing links in urban bicycle networks, using data from OpenStreetMap. In this procedure we first identify all possible gaps following a multiplex network approach, prioritize them according to a flow-based metric, decluster emerging gap clusters, and manually classify the types of gaps. We apply the IPDC procedure to Copenhagen and report the 105 top priority gaps. For evaluation, we compare these gaps with the city's most recent Cycle Path Prioritization Plan and find considerable overlaps. Our results show how network analysis with minimal data requirements can serve as a cost-efficient support tool for bicycle network planning. By taking into account the whole city network for consolidating urban bicycle infrastructure, our data-driven framework can complement localized, manual planning processes for more effective, city-wide decision-making.

physics.soc-ph↗

arXiv2021

u-cf2vec: Representation Learning for Personalized Algorithm Selection in Recommender Systems

Collaborative Filtering (CF) has become the standard approach to solve recommendation systems (RS) problems. Collaborative Filtering algorithms try to make predictions about interests of a user by collecting the personal interests from multiple users. There are multiple CF algorithms, each one of them with its own biases. It is the Machine Learning practitioner that has to choose the best algorithm for each task beforehand. In Recommender Systems, different algorithms have different performance for different users within the same dataset. Meta Learning (MtL) has been used to choose the best algorithm for a given problem. Meta Learning is usually applied to select algorithms for a whole dataset. Adapting it to select the to the algorithm for a single user in a RS involves several challenges. The most important is the design of the metafeatures which, in typical meta learning, characterize datasets while here, they must characterize a single user. This work presents a new meta-learning based framework named $μ$-cf2vec to select the best algorithm for each user. We propose using Representation Learning techniques to extract the metafeatures. Representation Learning tries to extract representations that can be reused in other learning tasks. In this work we also implement the framework using different RL techniques to evaluate which one can be more useful to solve this task. In the meta level, the meta learning model will use the metafeatures to extract knowledge that will be used to predict the best algorithm for each user. We evaluated an implementation of this framework using MovieLens 20M dataset. Our implementation achieved consistent gains in the meta level, however, in the base level we only achieved marginal gains.

cs.IR↗

arXiv2019

Are All Successful Communities Alike? Characterizing and Predicting the Success of Online Communities

The proliferation of online communities has created exciting opportunities to study the mechanisms that explain group success. While a growing body of research investigates community success through a single measure -- typically, the number of members -- we argue that there are multiple ways of measuring success. Here, we present a systematic study to understand the relations between these success definitions and test how well they can be predicted based on community properties and behaviors from the earliest period of a community's lifetime. We identify four success measures that are desirable for most communities: (i) growth in the number of members; (ii) retention of members; (iii) long term survival of the community; and (iv) volume of activities within the community. Surprisingly, we find that our measures do not exhibit very high correlations, suggesting that they capture different types of success. Additionally, we find that different success measures are predicted by different attributes of online communities, suggesting that success can be achieved through different behaviors. Our work sheds light on the basic understanding of what success represents in online communities and what predicts it. Our results suggest that success is multi-faceted and cannot be measured nor predicted by a single measurement. This insight has practical implications for the creation of new online communities and the design of platforms that facilitate such communities.

cs.SI↗

arXiv2018

cf2vec: Collaborative Filtering algorithm selection using graph distributed representations

Algorithm selection using Metalearning aims to find mappings between problem characteristics (i.e. metafeatures) with relative algorithm performance to predict the best algorithm(s) for new datasets. Therefore, it is of the utmost importance that the metafeatures used are informative. In Collaborative Filtering, recent research has created an extensive collection of such metafeatures. However, since these are created based on the practitioner's understanding of the problem, they may not capture the most relevant aspects necessary to properly characterize the problem. We propose to overcome this problem by taking advantage of Representation Learning, which is able to create an alternative problem characterizations by having the data guide the design of the representation instead of the practitioner's opinion. Our hypothesis states that such alternative representations can be used to replace standard metafeatures, hence hence leading to a more robust approach to Metalearning. We propose a novel procedure specially designed for Collaborative Filtering algorithm selection. The procedure models Collaborative Filtering as graphs and extracts distributed representations using graph2vec. Experimental results show that the proposed procedure creates representations that are competitive with state-of-the-art metafeatures, while requiring significantly less data and without virtually any human input.

cs.IR↗

arXiv2018

Algorithm Selection for Collaborative Filtering: the influence of graph metafeatures and multicriteria metatargets

To select the best algorithm for a new problem is an expensive and difficult task. However, there are automatic solutions to address this problem: using Metalearning, which takes advantage of problem characteristics (i.e. metafeatures), one is able to predict the relative performance of algorithms. In the Collaborative Filtering scope, recent works have proposed diverse metafeatures describing several dimensions of this problem. Despite interesting and effective findings, it is still unknown whether these are the most effective metafeatures. Hence, this work proposes a new set of graph metafeatures, which approach the Collaborative Filtering problem from a Graph Theory perspective. Furthermore, in order to understand whether metafeatures from multiple dimensions are a better fit, we investigate the effects of comprehensive metafeatures. These metafeatures are a selection of the best metafeatures from all existing Collaborative Filtering metafeatures. The impact of the most representative metafeatures is investigated in a controlled experimental setup. Another contribution we present is the use of a Pareto-Efficient ranking procedure to create multicriteria metatargets. These new rankings of algorithms, which take into account multiple evaluation measures, allow to explore the algorithm selection problem in a fairer and more detailed way. According to the experimental results, the graph metafeatures are a good alternative to related work metafeatures. However, the results have shown that the feature selection procedure used to create the comprehensive metafeatures is is not effective, since there is no gain in predictive performance. Finally, an extensive metaknowledge analysis was conducted to identify the most influential metafeatures.

cs.IR↗

arXiv2018

CF4CF: Recommending Collaborative Filtering algorithms using Collaborative Filtering

Automatic solutions which enable the selection of the best algorithms for a new problem are commonly found in the literature. One research area which has recently received considerable efforts is Collaborative Filtering. Existing work includes several approaches using Metalearning, which relate the characteristics of datasets with the performance of the algorithms. This work explores an alternative approach to tackle this problem. Since, in essence, both are recommendation problems, this work uses Collaborative Filtering algorithms to select Collaborative Filtering algorithms. Our approach integrates subsampling landmarkers, which are a data characterization approach commonly used in Metalearning, with a standard Collaborative Filtering method. The experimental results show that CF4CF competes with standard Metalearning strategies in the problem of Collaborative Filtering algorithm selection.

cs.IR↗