arXiv subjects

Swaprava Nath

Publications and source records attributed to Swaprava Nath.

At least 19 recordsLinked to original sources

arXiv2026

Fair Allocation under Conflict Constraints

We study the fair allocation of indivisible items subject to conflict constraints. In this framework, the items are represented as the vertices of a graph, with edges corresponding to conflicts between pairs of items. Each agent is assigned an independent set of items from the graph. Our goal is to achieve a fair and efficient allocation of these items. Fairness pertains to satisfying envy-freeness up to one item (EF1), while efficiency is defined by maximality, meaning that no unallocated item can be feasibly assigned to any agent. First, we explore the case of two agents. For monotone valuations, we show that a maximal EF1 allocation always exists on any graph. Our existence proof relies on a color-switching technique, which locally modifies a maximal allocation while preserving feasibility and restoring EF1. We further show that such allocations can be computed in pseudopolynomial time in general, and in polynomial time for additive valuations on arbitrary graphs, as well as for monotone valuations on interval and bipartite graphs. By contrast, once monotonicity is dropped, maximal EF1 allocations need not exist even for identical additive valuations, and deciding existence becomes NP-hard. Next, we consider the case with a general number of agents. Again, we arrive at a negative result: An EF1 and maximal allocation fails to exist even for three agents under identical monotone valuations, and determining the existence of such an allocation is NP-hard. On the positive side, we show that under identical non-monotone additive valuations on a path graph, an EF[1,1] and maximal allocation always exists. This result involves a novel application of the "cycle plus triangles" theorem.

cs.GT↗

arXiv2025

Likes, Budgets, and Equilibria: Designing Contests for Socially Optimal Advertising

Firms (businesses, service providers, entertainment organizations, political parties, etc.) advertise on social networks to draw people's attention and improve their awareness of the brands of the firms. In all such cases, the competitive nature of their engagements gives rise to a game where the firms need to decide how to distribute their budget over the agents on a network to maximize their brand's awareness. The firms (players) therefore need to optimize how much budget they should put on the vertices of the network so that the spread improves via direct (via advertisements or free promotional offers) and indirect marketing (words-of-mouth). We propose a two-timescale model of decisions where the communication between the vertices happen in a faster timescale and the strategy update of the firms happen in a slower timescale. We show that under fairly standard conditions, the best response dynamics of the firms converge to a pure strategy Nash equilibrium. However, such equilibria can be away from a socially optimal one. We provide a characterization of the contest success functions and provide examples for the designers of such contests (e.g., regulators, social network providers, etc.) such that the Nash equilibrium becomes unique and social welfare maximizing. Our experiments show that for realistic scenarios, such contest success functions perform fairly well.

cs.GT↗

arXiv2025

Sequential Solution Concepts in Cooperative Games with Generalized Characteristic Functions

Motivated by the fact that the worth of a coalition may depend on the order in which agents arrive, Nowak and Radzik (1994) (NR) introduced cooperative games with generalized characteristic functions. We study such temporal cooperative games (TCGs), where the worth function v is defined on sequences of agents {\pi} rather than sets S. This order sensitivity necessitates a re-examination of axioms for reward sharing. NR and subsequent work proposed several axioms; the resulting solution concepts are still inherently order-oblivious and closely tied to the Shapley value. In contrast, we focus on sequential solution concepts that explicitly depend on the realized order {\pi}. We study reward-sharing mechanisms satisfying incentive for optimal arrival (I4OA), which promotes orders maximizing total worth; online individual rationality (OIR), which ensures agents are not harmed by later arrivals; and sequential efficiency (SE), which requires that the worth of any sequence is fully distributed among its agents. These axioms are intrinsic to TCGs, and we characterize a class of reward-sharing mechanisms uniquely determined by them. The classical Shapley value does not directly extend to this setting. We therefore construct natural Shapley analogs in two worlds: a sequential world, where rewards are defined for each sequence agent pair, and an extended world, where rewards are defined per agent, consistent with the NR framework. In both cases, the axioms of efficiency, additivity, and null player uniquely characterize the corresponding Shapley analogs. But, these Shapley analogs are disjoint from the class of solutions satisfying the sequential axioms, even for convex and simple TCGs.

cs.GT↗

arXiv2025

Incentivize Contribution and Learn Parameters Too: Federated Learning with Strategic Data Owners

Classical federated learning (FL) assumes that the clients have a limited amount of noisy data with which they voluntarily participate and contribute towards learning a global, more accurate model in a principled manner. The learning happens in a distributed fashion without sharing the data with the center. However, these methods do not consider the incentive of an agent for participating and contributing to the process, given that data collection and running a distributed algorithm is costly for the clients. The question of rationality of contribution has been asked recently in the literature and some results exist that consider this problem. This paper addresses the question of simultaneous parameter learning and incentivizing contribution in a truthful manner, which distinguishes it from the extant literature. Our first mechanism incentivizes each client to contribute to the FL process at a Nash equilibrium and simultaneously learn the model parameters. We also ensure that agents are incentivized to truthfully reveal information in the intermediate stages of the algorithm. However, this equilibrium outcome can be away from the optimal, where clients contribute with their full data and the algorithm learns the optimal parameters. We propose a second mechanism that enables the full data contribution along with optimal parameter learning. Large scale experiments with real (federated) datasets (CIFAR-10, FEMNIST, and Twitter) show that these algorithms converge quite fast in practice, yield good welfare guarantees and better model performance for all agents.

cs.GT↗

arXiv2024

A Gale-Shapley View of Unique Stable Marriages

Stable marriage of a two-sided market with unit demand is a classic problem that arises in many real-world scenarios. In addition, a unique stable marriage in this market simplifies a host of downstream desiderata. In this paper, we explore a new set of sufficient conditions for unique stable matching (USM) under this setup. Unlike other approaches that also address this question using the structure of preference profiles, we use an algorithmic viewpoint and investigate if this question can be answered using the lens of the deferred acceptance (DA) algorithm (Gale and Shapley, 1962). Our results yield a set of sufficient conditions for USM (viz., MaxProp and MaxRou) and show that these are disjoint from the previously known sufficiency conditions like sequential preference and no crossing. We also provide a characterization of MaxProp that makes it efficiently verifiable, and shows the gap between MaxProp and the entire USM class. These results give a more detailed view of the sub-structures of the USM class.

cs.GT↗

arXiv2024

Distilling Opinions at Scale: Incremental Opinion Summarization using XL-OPSUMM

Opinion summarization in e-commerce encapsulates the collective views of numerous users about a product based on their reviews. Typically, a product on an e-commerce platform has thousands of reviews, each review comprising around 10-15 words. While Large Language Models (LLMs) have shown proficiency in summarization tasks, they struggle to handle such a large volume of reviews due to context limitations. To mitigate, we propose a scalable framework called Xl-OpSumm that generates summaries incrementally. However, the existing test set, AMASUM has only 560 reviews per product on average. Due to the lack of a test set with thousands of reviews, we created a new test set called Xl-Flipkart by gathering data from the Flipkart website and generating summaries using GPT-4. Through various automatic evaluations and extensive analysis, we evaluated the framework's efficiency on two datasets, AMASUM and Xl-Flipkart. Experimental results show that our framework, Xl-OpSumm powered by Llama-3-8B-8k, achieves an average ROUGE-1 F1 gain of 4.38% and a ROUGE-L F1 gain of 3.70% over the next best-performing model.

cs.CL↗

arXiv2024

One Prompt To Rule Them All: LLMs for Opinion Summary Evaluation

Evaluation of opinion summaries using conventional reference-based metrics rarely provides a holistic evaluation and has been shown to have a relatively low correlation with human judgments. Recent studies suggest using Large Language Models (LLMs) as reference-free metrics for NLG evaluation, however, they remain unexplored for opinion summary evaluation. Moreover, limited opinion summary evaluation datasets inhibit progress. To address this, we release the SUMMEVAL-OP dataset covering 7 dimensions related to the evaluation of opinion summaries: fluency, coherence, relevance, faithfulness, aspect coverage, sentiment consistency, and specificity. We investigate Op-I-Prompt a dimension-independent prompt, and Op-Prompts, a dimension-dependent set of prompts for opinion summary evaluation. Experiments indicate that Op-I-Prompt emerges as a good alternative for evaluating opinion summaries achieving an average Spearman correlation of 0.70 with humans, outperforming all previous approaches. To the best of our knowledge, we are the first to investigate LLMs as evaluators on both closed-source and open-source models in the opinion summarization domain.

cs.CL↗

arXiv2024

Product Description and QA Assisted Self-Supervised Opinion Summarization

In e-commerce, opinion summarization is the process of summarizing the consensus opinions found in product reviews. However, the potential of additional sources such as product description and question-answers (QA) has been considered less often. Moreover, the absence of any supervised training data makes this task challenging. To address this, we propose a novel synthetic dataset creation (SDC) strategy that leverages information from reviews as well as additional sources for selecting one of the reviews as a pseudo-summary to enable supervised training. Our Multi-Encoder Decoder framework for Opinion Summarization (MEDOS) employs a separate encoder for each source, enabling effective selection of information while generating the summary. For evaluation, due to the unavailability of test sets with additional sources, we extend the Amazon, Oposum+, and Flipkart test sets and leverage ChatGPT to annotate summaries. Experiments across nine test sets demonstrate that the combination of our SDC approach and MEDOS model achieves on average a 14.5% improvement in ROUGE-1 F1 over the SOTA. Moreover, comparative analysis underlines the significance of incorporating additional sources for generating more informative summaries. Human evaluations further indicate that MEDOS scores relatively higher in coherence and fluency with 0.41 and 0.5 (-1 to 1) respectively, compared to existing models. To the best of our knowledge, we are the first to generate opinion summaries leveraging additional sources in a self-supervised setting.

cs.CL↗

arXiv2024

Fair Interval Scheduling of Indivisible Chores

We study the problem of fairly assigning a set of discrete tasks (or chores) among a set of agents with additive valuations. Each chore is associated with a start and finish time, and each agent can perform at most one chore at any given time. The goal is to find a fair and efficient schedule of the chores, where fairness pertains to satisfying envy-freeness up to one chore (EF1) and efficiency pertains to maximality (i.e., no unallocated chore can be feasibly assigned to any agent). Our main result is a polynomial-time algorithm for computing an EF1 and maximal schedule for two agents under monotone valuations when the conflict constraints constitute an arbitrary interval graph. The algorithm uses a coloring technique in interval graphs that may be of independent interest. For an arbitrary number of agents with identical additive valuations, we show the existence of an EF1 and maximal schedule when the constraints constitute a path graph. This result uses a reduction to the ``cycle-plus-triangles'' theorem. Using different techniques, we provide an efficient algorithm for finding such a schedule when there are four or more agents and the valuations are further assumed to be dichotomous. We also show that stronger fairness and efficiency properties, including envy-freeness up to any chore (EFX) along with maximality and EF1 along with Pareto optimality, cannot be achieved.

cs.GT↗

arXiv2023

Disentangling Societal Inequality from Model Biases: Gender Inequality in Divorce Court Proceedings

Divorce is the legal dissolution of a marriage by a court. Since this is usually an unpleasant outcome of a marital union, each party may have reasons to call the decision to quit which is generally documented in detail in the court proceedings. Via a substantial corpus of 17,306 court proceedings, this paper investigates gender inequality through the lens of divorce court proceedings. While emerging data sources (e.g., public court records) on sensitive societal issues hold promise in aiding social science research, biases present in cutting-edge natural language processing (NLP) methods may interfere with or affect such studies. We thus require a thorough analysis of potential gaps and limitations present in extant NLP resources. In this paper, on the methodological side, we demonstrate that existing NLP resources required several non-trivial modifications to quantify societal inequalities. On the substantive side, we find that while a large number of court cases perhaps suggest changing norms in India where women are increasingly challenging patriarchy, AI-powered analyses of these court proceedings indicate striking gender inequality with women often subjected to domestic violence.

cs.CY↗

arXiv2023

Optimal Referral Auction Design

The auction of a single indivisible item is one of the most celebrated problems in mechanism design with transfers. Despite its simplicity, it provides arguably the cleanest and most insightful results in the literature. When the information that the auction is running is available to every participant, Myerson [20] provided a seminal result to characterize the incentive-compatible auctions along with revenue optimality. However, such a result does not hold in an auction on a network, where the information of the auction is spread via the agents, and they need incentives to forward the information. In recent times, a few auctions (e.g., [13, 18]) were designed that appropriately incentivized the intermediate nodes on the network to promulgate the information to potentially more valuable bidders. In this paper, we provide a Myerson-like characterization of incentive-compatible auctions on a network and show that the currently known auctions fall within this class of randomized auctions. We then consider a special class called the referral auctions that are inspired by the multi-level marketing mechanisms [1, 6, 7] and obtain the structure of a revenue optimal referral auction for i.i.d. bidders. Through experiments, we show that even for non-i.i.d. bidders there exist auctions following this characterization that can provide a higher revenue than the currently known auctions on networks.

cs.GT↗

arXiv2021

Social Distancing via Social Scheduling

Motivated by the need of {\em social distancing} during a pandemic, we consider an approach to schedule the visitors of a facility (e.g., a general store). Our algorithms take input from the citizens and schedule the store's discrete time-slots based on their importance to visit the facility. Naturally, the formulation applies to several similar problems. We consider {\em indivisible} job requests that take single or multiple slots to complete. The salient properties of our approach are: it (a)~ensures social distancing by ensuring a maximum population in a given time-slot at the facility, (b)~aims to prioritize individuals based on the importance of the jobs, (c)~maintains truthfulness of the reported importance by adding a {\em cooling-off} period after their allocated time-slot, during which the individual cannot re-access the same facility, (d)~guarantees voluntary participation of the citizens, and yet (e)~is computationally tractable. The mechanisms we propose are prior-free. We show that the problem becomes NP-complete for indivisible multi-slot demands, and provide a polynomial-time mechanism that is truthful, individually rational, and approximately optimal. Experiments with data collected from a store show that visitors with more important (single-slot) jobs are allocated more preferred slots, which comes at the cost of a longer cooling-off period and significantly reduces social congestion. For the multi-slot jobs, our mechanism yields reasonable approximation while reducing the computation time significantly.

cs.GT↗

arXiv2021

Removing Bias and Incentivizing Precision in Peer-grading

We study peer-grading with competitive graders who enjoy a higher utility when their peers get lower scores. We propose a new mechanism, PEQA, that incentivizes such graders through a score-assignment rule which aggregates the final score from multiple peer-evaluations, and a grading performance score that rewards performance in the peer-grading exercise. PEQA makes grader-bias irrelevant. Additionally, under PEQA, a peer-grader's utility increases monotonically with the reliability of her grading, irrespective of her competitiveness and how her co-graders act. In a reasonably general class of score assignment rules, PEQA uniquely satisfies this utility- reliability monotonicity. When grading is costly and costs are private information, a modified version of PEQA implements the socially optimal effort choices in an equilibrium of the peer-evaluation game. Data from our classroom experiments confirm our theoretical assumptions and show that PEQA outperforms the popular median mechanism.

cs.GT↗

arXiv2021

Egalitarian and Congestion Aware Truthful Airport Slot Allocation Mechanism

We propose a mechanism to allocate slots fairly at congested airports. This mechanism: (a) ensures that the slots are allocated according to the true valuations of airlines, (b) provides fair opportunities to the flights connecting remote cities to large airports, and (c) controls the number of flights in each slot to minimize congestion. The mechanism draws inspiration from economic theory. It allocates the slots based on an affine maximizer allocation rule and charges payments to the airlines such that they are incentivized to reveal their true valuations. The allocation also optimizes the occupancy of every slot to keep them as uncongested as possible. The formulation solves an optimal integral solution in strongly polynomial time. We conduct experiments on the data collected from two major airports in India. We also compare our results with existing allocations and also with the allocations based on the International Air Transport Association (IATA) guidelines. The computational results show that the social utility generated using our mechanism is 20-30% higher than IATA and current allocations.

cs.GT↗

arXiv2021

TruthBot: An Automated Conversational Tool for Intent Learning, Curated Information Presenting, and Fake News Alerting

We present TruthBot, an all-in-one multilingual conversational chatbot designed for seeking truth (trustworthy and verified information) on specific topics. It helps users to obtain information specific to certain topics, fact-check information, and get recent news. The chatbot learns the intent of a query by training a deep neural network from the data of the previous intents and responds appropriately when it classifies the intent in one of the classes above. Each class is implemented as a separate module that uses either its own curated knowledge-base or searches the web to obtain the correct information. The topic of the chatbot is currently set to COVID-19. However, the bot can be easily customized to any topic-specific responses. Our experimental results show that each module performs significantly better than its closest competitor, which is verified both quantitatively and through several user-based surveys in multiple languages. TruthBot has been deployed in June 2020 and is currently running.

cs.SI↗

arXiv2020

SkillCheck: An Incentive-based Certification System using Blockchains

Skill verification is a central problem in workforce hiring. Companies and academia often face the difficulty of ascertaining the skills of an applicant since the certifications of the skills claimed by a candidate are generally not immediately verifiable and costly to test. Blockchains have been proposed in the literature for skill verification and tamper-proof information storage in a decentralized manner. However, most of these approaches deal with storing the certificates issued by traditional universities on the blockchain. Among the few techniques that consider the certification procedure itself, questions like (a) scalability with limited staff, (b) uniformity of grades over multiple evaluators, or (c) honest effort extraction from the evaluators are usually not addressed. We propose a blockchain-based platform named SkillCheck, which considers the questions above, and ensure several desirable properties. The platform incentivizes effort in grading via payments with tokens which it generates from the payments of the users of the platform, e.g., the recruiters and test-takers. We provide a detailed description of the design of the platform along with the provable properties of the algorithm.

cs.CR↗

arXiv2020

Truthful ownership transfer with expert advice: Blending mechanism design with and without money

When a company undergoes a merger or transfers its ownership, the existing governing body has an opinion on which buyer should take over as the new owner. Similar situations occur while assigning the host of big sports tournaments, like the World Cup or the Olympics. In all these settings, the values of the external bidders are as important as the opinions of the internal experts. Motivated by such scenarios, we consider a social welfare maximizing approach to design and analyze truthful mechanisms in {\em hybrid social choice} settings, where payments can be imposed to the bidders, but not to the experts. Since this problem is a combination of mechanism design with and without monetary transfers, classical solutions like VCG cannot be applied, making this a novel mechanism design problem. We consider the simple but fundamental scenario with one expert and two bidders, and provide tight approximation guarantees of the optimal social welfare. We distinguish between mechanisms that use ordinal and cardinal information, as well as between mechanisms that base their decisions on one of the two sides (either the bidders or the expert) or both. Our analysis shows that the cardinal setting is quite rich and admits several non-trivial randomized truthful mechanisms, and also allows for closer-to-optimal welfare guarantees.

cs.GT↗

arXiv2019

SPARCAS: A Decentralized, Truthful Multi-Agent Collision-free Path Finding Mechanism

We propose a decentralized collision-avoidance mechanism for a group of independently controlled robots moving on a shared workspace. Existing algorithms achieve multi-robot collision avoidance either (a) in a centralized setting, or (b) in a decentralized setting with collaborative robots. We focus on the setting with competitive robots in a decentralized environment, where robots may strategically reveal their information to get prioritized. We propose the mechanism SPARCAS in this setting that, using principles of mechanism design, ensures truthful revelation of the robots' private information and provides locally efficient movement of the robots. It is free from collisions and deadlocks, and handles a dynamic arrival of robots. In practice, this mechanism scales well for a large number of robots where the optimal collision-avoiding path-finding algorithm (M*) does not scale. Yet, SPARCAS does not compromise the path optimality too much. Our mechanism prioritizes the robots in the order of their `true' higher needs, but for a higher payment. It uses monetary transfers which is small enough compared to the value received by the robots.

cs.RO↗