arXiv subjects

Sourish Dasgupta

Publications and source records attributed to Sourish Dasgupta.

14 recordsLinked to original sources

arXiv2026

HalluPrism: When Multimodal Uncertainty Should Diagnose, Not Decide

Multimodal Large Language Models (MLLMs) can assign similar confidence to answers that fail for different reasons. We propose HalluPrism, a behavioral diagnostic that re-runs an answer after visual degradation, blank-image replacement, and grounding or relation checks. These targeted probes yield a signature over visual-perturbation sensitivity (V ), image-removal confidence retention (L), and grounding/relation-probe instability (A). Across 58K+ examples from four benchmarks and four MLLMs, image-removal confidence retention is most prevalent, while grounding/relation-probe instability better separates failure families. Only 18 of 48 source-target checks are diagonally aligned, so the coordinates should be interpreted jointly rather than as independent causal sources. With the dataset fixed, the joint signature improves failure-family AUROC from 0.634 to 0.769 on HallusionBench and from 0.707 to 0.817 on VizWiz, with smaller gains on POPE and VSR. In pooled XGBoost analysis, AUROC rises from 0.78 with scalar confidence to 0.95 with (V, L, A) and 0.97 when confidence is added. The same signature does not automatically improve correctness ranking. The three tested direct scalarizations can harm it. These results separate failure diagnosis from abstention scoring: multimodal uncertainty should characterize failure structure before it is used to decide whether to abstain or correct.

cs.LG↗

arXiv2026

Polaris: Coupled Orbital Polar Embeddings for Hierarchical Concept Learning

Real-world knowledge is often organized as hierarchies such as product taxonomies, medical ontologies, and label trees, yet learning hierarchical representations is challenging due to asymmetric structure and noisy semantics. We introduce Polaris, a polar hyperspherical embedding framework that separates semanticity from hierarchy using angular geometry and radius, enabling the learning of meaning and structure without interference. To map latent representation onto the sphere, we project it to the tangent space at the north pole, apply the exponential map, and learn unit-norm representations using spherical linear layers. Polaris then combines robust local constraints, global regularization that prevents geometric collapse, and uncertainty-aware asymmetric objectives that encourage directional containment. At inference time, Polaris uses structure-guided retrieval to efficiently narrow down candidate parents before final ranking. We evaluate Polaris on different settings of taxonomy expansion - spanning trees, multi-parent DAGs, and multimodal hierarchies, showing consistent improvements of up to ~19 points in top-K retrieval and up to ~60% reduction in mean rank over fourteen strong baselines.

cs.LG↗

arXiv2025

Diversity Augmentation of Dynamic User Preference Data for Boosting Personalized Text Summarizers

Document summarization enables efficient extraction of user-relevant content but is inherently shaped by individual subjectivity, making it challenging to identify subjective salient information in multifaceted documents. This complexity underscores the necessity for personalized summarization. However, training models for personalized summarization has so far been challenging, particularly because diverse training data containing both user preference history (i.e., click-skip trajectory) and expected (gold-reference) summaries are scarce. The MS/CAS PENS dataset is a valuable resource but includes only preference history without target summaries, preventing end-to-end supervised learning, and its limited topic-transition diversity further restricts generalization. To address this, we propose $\mathrm{PerAugy}$, a novel cross-trajectory shuffling and summary-content perturbation based data augmentation technique that significantly boosts the accuracy of four state-of-the-art baseline (SOTA) user-encoders commonly used in personalized summarization frameworks (best result: $\text{0.132}$$\uparrow$ w.r.t AUC). We select two such SOTA summarizer frameworks as baselines and observe that when augmented with their corresponding improved user-encoders, they consistently show an increase in personalization (avg. boost: $\text{61.2\%}\uparrow$ w.r.t. PSE-SU4 metric). As a post-hoc analysis of the role of induced diversity in the augmented dataset by \peraugy, we introduce three dataset diversity metrics -- $\mathrm{TP}$, $\mathrm{RTC}$, and \degreed\ to quantify the induced diversity. We find that $\mathrm{TP}$ and $\mathrm{DegreeD}$ strongly correlate with user-encoder performance on the PerAugy-generated dataset across all accuracy metrics, indicating that increased dataset diversity is a key factor driving performance gains.

cs.CL↗

arXiv2024

PerSEval: Assessing Personalization in Text Summarizers

Personalized summarization models cater to individuals' subjective understanding of saliency, as represented by their reading history and current topics of attention. Existing personalized text summarizers are primarily evaluated based on accuracy measures such as BLEU, ROUGE, and METEOR. However, a recent study argued that accuracy measures are inadequate for evaluating the degree of personalization of these models and proposed EGISES, the first metric to evaluate personalized text summaries. It was suggested that accuracy is a separate aspect and should be evaluated standalone. In this paper, we challenge the necessity of an accuracy leaderboard, suggesting that relying on accuracy-based aggregated results might lead to misleading conclusions. To support this, we delve deeper into EGISES, demonstrating both theoretically and empirically that it measures the degree of responsiveness, a necessary but not sufficient condition for degree-of-personalization. We subsequently propose PerSEval, a novel measure that satisfies the required sufficiency condition. Based on the benchmarking of ten SOTA summarization models on the PENS dataset, we empirically establish that -- (i) PerSEval is reliable w.r.t human-judgment correlation (Pearson's r = 0.73; Spearman's $ρ$ = 0.62; Kendall's $τ$ = 0.42), (ii) PerSEval has high rank-stability, (iii) PerSEval as a rank-measure is not entailed by EGISES-based ranking, and (iv) PerSEval can be a standalone rank-measure without the need of any aggregated ranking.

cs.CL↗

arXiv2024

Are Large Language Models In-Context Personalized Summarizers? Get an iCOPERNICUS Test Done!

Large Language Models (LLMs) have succeeded considerably in In-Context-Learning (ICL) based summarization. However, saliency is subject to the users' specific preference histories. Hence, we need reliable In-Context Personalization Learning (ICPL) capabilities within such LLMs. For any arbitrary LLM to exhibit ICPL, it needs to have the ability to discern contrast in user profiles. A recent study proposed a measure for degree-of-personalization called EGISES for the first time. EGISES measures a model's responsiveness to user profile differences. However, it cannot test if a model utilizes all three types of cues provided in ICPL prompts: (i) example summaries, (ii) user's reading histories, and (iii) contrast in user profiles. To address this, we propose the iCOPERNICUS framework, a novel In-COntext PERsonalization learNIng sCrUtiny of Summarization capability in LLMs that uses EGISES as a comparative measure. As a case-study, we evaluate 17 state-of-the-art LLMs based on their reported ICL performances and observe that 15 models' ICPL degrades (min: 1.6%; max: 3.6%) when probed with richer prompts, thereby showing lack of true ICPL.

cs.CL↗

arXiv2023

Inline Citation Classification using Peripheral Context and Time-evolving Augmentation

Citation plays a pivotal role in determining the associations among research articles. It portrays essential information in indicative, supportive, or contrastive studies. The task of inline citation classification aids in extrapolating these relationships; However, existing studies are still immature and demand further scrutiny. Current datasets and methods used for inline citation classification only use citation-marked sentences constraining the model to turn a blind eye to domain knowledge and neighboring contextual sentences. In this paper, we propose a new dataset, named 3Cext, which along with the cited sentences, provides discourse information using the vicinal sentences to analyze the contrasting and entailing relationships as well as domain information. We propose PeriCite, a Transformer-based deep neural network that fuses peripheral sentences and domain knowledge. Our model achieves the state-of-the-art on the 3Cext dataset by +0.09 F1 against the best baseline. We conduct extensive ablations to analyze the efficacy of the proposed dataset and model fusion methods.

cs.CL↗

arXiv2018

Formal Ontology Learning from English IS-A Sentences

Ontology learning (OL) is the process of automatically generating an ontological knowledge base from a plain text document. In this paper, we propose a new ontology learning approach and tool, called DLOL, which generates a knowledge base in the description logic (DL) SHOQ(D) from a collection of factual non-negative IS-A sentences in English. We provide extensive experimental results on the accuracy of DLOL, giving experimental comparisons to three state-of-the-art existing OL tools, namely Text2Onto, FRED, and LExO. Here, we use the standard OL accuracy measure, called lexical accuracy, and a novel OL accuracy measure, called instance-based inference model. In our experimental results, DLOL turns out to be about 21% and 46%, respectively, better than the best of the other three approaches.

cs.AI↗

arXiv2017

SimDoc: Topic Sequence Alignment based Document Similarity Framework

Document similarity is the problem of estimating the degree to which a given pair of documents has similar semantic content. An accurate document similarity measure can improve several enterprise relevant tasks such as document clustering, text mining, and question-answering. In this paper, we show that a document's thematic flow, which is often disregarded by bag-of-word techniques, is pivotal in estimating their similarity. To this end, we propose a novel semantic document similarity framework, called SimDoc. We model documents as topic-sequences, where topics represent latent generative clusters of related words. Then, we use a sequence alignment algorithm to estimate their semantic similarity. We further conceptualize a novel mechanism to compute topic-topic similarity to fine tune our system. In our experiments, we show that SimDoc outperforms many contemporary bag-of-words techniques in accurately computing document similarity, and on practical applications such as document clustering.

cs.CL↗

arXiv2017

User Profile Based Research Paper Recommendation

We design a recommender system for research papers based on topic-modeling. The users feedback to the results is used to make the results more relevant the next time they fire a query. The user's needs are understood by observing the change in the themes that the user shows a preference for over time.

cs.IR↗

arXiv2016

Formal Ontology Learning on Factual IS-A Corpus in English using Description Logics

Ontology Learning (OL) is the computational task of generating a knowledge base in the form of an ontology given an unstructured corpus whose content is in natural language (NL). Several works can be found in this area most of which are limited to statistical and lexico-syntactic pattern matching based techniques Light-Weight OL. These techniques do not lead to very accurate learning mostly because of several linguistic nuances in NL. Formal OL is an alternative (less explored) methodology were deep linguistics analysis is made using theory and tools found in computational linguistics to generate formal axioms and definitions instead simply inducing a taxonomy. In this paper we propose "Description Logic (DL)" based formal OL framework for learning factual IS-A type sentences in English. We claim that semantic construction of IS-A sentences is non trivial. Hence, we also claim that such sentences requires special studies in the context of OL before any truly formal OL can be proposed. We introduce a learner tool, called DLOL_IS-A, that generated such ontologies in the owl format. We have adopted "Gold Standard" based OL evaluation on IS-A rich WCL v.1.1 dataset and our own Community representative IS-A dataset. We observed significant improvement of DLOL_IS-A when compared to the light-weight OL tool Text2Onto and formal OL tool FRED.

cs.CL↗

arXiv2015

BitSim: An Algebraic Similarity Measure for Description Logics Concepts

In this paper, we propose an algebraic similarity measure σBS (BS stands for BitSim) for assigning semantic similarity score to concept definitions in ALCH+ an expressive fragment of Description Logics (DL). We define an algebraic interpretation function, I_B, that maps a concept definition to a unique string (ω_B) called bit-code) over an alphabet Σ_B of 11 symbols belonging to L_B - the language over P B. IB has semantic correspondence with conventional model-theoretic interpretation of DL. We then define σ_BS on L_B. A detailed analysis of I_B and σ_BS has been given.

cs.AI↗

arXiv2013

Description Logics based Formalization of Wh-Queries

The problem of Natural Language Query Formalization (NLQF) is to translate a given user query in natural language (NL) into a formal language so that the semantic interpretation has equivalence with the NL interpretation. Formalization of NL queries enables logic based reasoning during information retrieval, database query, question-answering, etc. Formalization also helps in Web query normalization and indexing, query intent analysis, etc. In this paper we are proposing a Description Logics based formal methodology for wh-query intent (also called desire) identification and corresponding formal translation. We evaluated the scalability of our proposed formalism using Microsoft Encarta 98 query dataset and OWL-S TC v.4.0 dataset.

cs.CL↗

arXiv2013

STC: Semantic Taxonomical Clustering for Service Category Learning

Service discovery is one of the key problems that has been widely researched in the area of Service Oriented Architecture (SOA) based systems. Service category learning is a technique for efficiently facilitating service discovery. Most approaches for service category learning are based on suitable similarity distance measures using thresholds. Threshold selection is essentially difficult and often leads to unsatisfactory accuracy. In this paper, we have proposed a self-organizing based clustering algorithm called Semantic Taxonomical Clustering (STC) for taxonomically organizing services with self-organizing information and knowledge. We have tested the STC algorithm on both randomly generated data and the standard OWL-S TC dataset. We have observed promising results both in terms of classification accuracy and runtime performance compared to existing approaches.

cs.SE↗

arXiv2013

DLOLIS-A: Description Logic based Text Ontology Learning

Ontology Learning has been the subject of intensive study for the past decade. Researchers in this field have been motivated by the possibility of automatically building a knowledge base on top of text documents so as to support reasoning based knowledge extraction. While most works in this field have been primarily statistical (known as light-weight Ontology Learning) not much attempt has been made in axiomatic Ontology Learning (called heavy-weight Ontology Learning) from Natural Language text documents. Heavy-weight Ontology Learning supports more precise formal logic-based reasoning when compared to statistical ontology learning. In this paper we have proposed a sound Ontology Learning tool DLOL_(IS-A) that maps English language IS-A sentences into their equivalent Description Logic (DL) expressions in order to automatically generate a consistent pair of T-box and A-box thereby forming both regular (definitional form) and generalized (axiomatic form) DL ontology. The current scope of the paper is strictly limited to IS-A sentences that exclude the possible structures of: (i) implicative IS-A sentences, and (ii) "Wh" IS-A questions. Other linguistic nuances that arise out of pragmatics and epistemic of IS-A sentences are beyond the scope of this present work. We have adopted Gold Standard based Ontology Learning evaluation on chosen IS-A rich Wikipedia documents.

cs.AI↗