Dissertation http://dspace.dtu.ac.in:8080/jspui/handle/repository/20350 2026-07-25T11:53:00Z 2026-07-25T11:53:00Z PRAKASH, KRISHNA Meena, Shweta (SUPERVISOR) http://dspace.dtu.ac.in:8080/jspui/handle/repository/23024 2026-07-06T09:18:24Z 2026-05-01T00:00:00Z

Title: ASTRUCTURAL AND STATEFUL APPROACH TO RETRIEVAL AUGMENTED GENERATION FOR CONTEXTUAL REASONING INKNOWLEDGE-INTENSIVE SYSTEMS Authors: PRAKASH, KRISHNA; Meena, Shweta (SUPERVISOR) Abstract: Systems that are built around large language models(LLMs) have to answer questions whose answers are provided by the model itself. Its accuracy degrades on hierarchical collections, which are revised over time, and that demand exact citations. This thesis develops a framework for such collections in two parts. The first part is a structural pipeline that respects the hierarchy of the source documents at every stage. It splits text into sections and clause boundaries, attaches citation-grade metadata to each chunk, runs a dense neural retriever and a sparse term-matching retriever in parallel, fuses their ranked lists, and reranks with A cross-encoder that uses the metadata as an extra signal, and constrains the language model to a fixed output format. The second part is a stateful agentic layer that wraps a lightweight controller around the structural pipeline. The controller maintains a working memory across retrieval calls, reflecting on the evidence collected so far, decides whether the evidence is sufficient, and issues a refined query when it is not. The structural pipeline is implemented and evaluated on a corpus of central Indian statutes against a benchmark of two hundred test queries whose ground-truth applicable provisions were prepared by a practising domain expert. The pipeline identifies the primary applicable citation in 94.0% of test queries (188 of 200; 95% Wilson interval [89.8%, 96.5%]) and returns the full applicable citation set in 82.0% (164 of 200; 95% Wilson interval [76.1%, 86.7%]). Every neural component is used without domain-specific fine-tuning, so the reported numbers reflect the pipeline’s behaviour in the absence of any domain-adaptation effect. The stateful agentic controller is presented as a fully specified design, composed from methods already published in the agentic and self-reflective retrieval literature; its end-to-end empirical evaluation is left for subsequent work.

2026-05-01T00:00:00Z KALIA, NITESH Singh, Priya (SUPERVISOR) http://dspace.dtu.ac.in:8080/jspui/handle/repository/23011 2026-07-06T09:16:39Z 2026-05-01T00:00:00Z

Title: LARGE LANGUAGE MODEL DRIVEN AGENTIC FRAME WORK FOR ADAPTIVE RETRIEVAL AND MULTI-HOP REASONING Authors: KALIA, NITESH; Singh, Priya (SUPERVISOR) Abstract: Retrieval Augmented Generation (RAG)hasestablished itself as a foundational paradigm for grounding the outputs of Large Language Models in externally retrieved factual ev idence, substantially reducing the hallucination that arises when parametric models are queried on knowledge beyond their training distribution. However, the dominant de ployment pattern for RAG remains a fixed single pass architecture in which a query is submitted once to a retrieval index, a static document set is returned, and an answer is generated from that set without any mechanism for self assessment, iterative refine ment, or adaptive re-retrieval. While this architecture is adequate for simple factoid questions whose complete answer can be recovered from a single retrieved document, it is fundamentally incapable of handling multi-hop questions, which are queries that require sequential reasoning across two or more documents where the answer to one intermediate reasoning step serves as the necessary context for formulating the next retrieval query. Failure mode is well characterised and systematic: If the single re trieved set is not complete (missing bridging facts), the generator either fabricates a plausible sounding but incorrect answer, or fails to complete an answer at all, without mechanisms to seek the missing evidence. This thesis directly tackles the above fault by designing, implementing, and eval uating thoroughly a Large Language Model (LLM) Driven Agentic Framework for Adaptive Retrieval and Multi-Hop Reasoning (MR). Three cooperative agentic mech anisms are introduced to the retrieval loop, adding to the passive fixed pipeline RAG architecture and making it an active, self-aware, evidence seeking process. The first is a question classifier that allows to classify an input query as simple or compositional and processes it through the right retrieval path, without causing the degradation of the accuracy that would result if complex decomposition strategies were applied to simple questions. The second is an iterative query rewriter that, at each retrieval step, explic itly identifies the specific information missing from the accumulated evidence set and constructs a targeted sub-query to retrieve it, appending only new documents to pre vent redundant context growth and terminating when no new evidence is returned. The third is a dual-signal sufficiency checker that combines a dense retrieval cosine simi larity threshold with an LLM based answerability judgement, requiring both signals to be satisfied simultaneously before answer generation is initiated, thereby reducing the probability of generating from insufficient evidence. The framework is evaluated on one hundred questions from the HotpotQA distrac tor benchmark, a dataset specifically designed to test multi-hop retrieval and reasoning under realistic retrieval noise conditions, achieving 46.0% Exact Match and 58.66% F1. These results represent absolute improvements of twenty four and sixteen per centage points respectively over a Vanilla RAG baseline, corresponding to a relative iv improvement of 109% in Exact Match, and outperform all three fixed strategy base lines compared. A structured ablation study establishes that all three components are necessary and complementary, with query rewriting identified as the single most crit ical component whose removal causes a ten percentage point Exact Match drop and a twenty five percentage point Retrieval Recall collapse. A comprehensive failure analy sis demonstrates that the framework successfully shifts the dominant error mode from retrieval failure to reasoning failure, directly redirecting future research effort toward generation quality rather than retrieval design. Together, these findings establish the proposed framework as a principled and empirically validated solution to the multi-hop retrieval and reasoning challenge that is practically deployable without any annotated reasoning supervision or model fine-tuning.

2026-05-01T00:00:00Z SINGH, VIBHOR http://dspace.dtu.ac.in:8080/jspui/handle/repository/22058 2025-08-08T05:18:16Z 2025-05-01T00:00:00Z

Title: A COMPARATIVE STUDY ON MACHINE LEARNING BASED STOCK PREDICTION BY INCORPORATING SENTIMENT ANALYSIS USING FINBERT Authors: SINGH, VIBHOR Abstract: Swift evolution of ML techniques, the financial forecasting scenario have been transformed tremendously. The domain of stock market prediction is one of the prominent applications of ML considering its inherent complexity and economic implication. Time series and statistical model was conventional cornerstone of market analysis, but they have limited potential for capturing intricate pattern. NLP sentiment analysis played a game changing role for more efficient stock prediction. NLP taps into huge pool of textual data of different sources, comprising of news and social media outlets. The collective mood and opinion of market participants can now be harness using power of NLP sentiment analysis which is to be fed to predictive model for better forecasting. Fusion of sentiment derived insights with ML algorithms presents a substantial leap which not only surges the predictive power of existing models but also provide nuanced understanding of the psychology of market movements driving factors. Consequently, financial industry witnessing a paradigm shift for the anticipation of stock prices fluctuations, with the support of AI driven sentiment analysis This paper presents a machine learning-based stock prediction model that integrates sentiment analysis using FinBERT, it is a specialized model for financial sentiment analysis that uses BERT. This study focuses on enhancing financial stock forecasting by adding investors sentiment data with conventional stock price data. This study takes into consideration traditional time series model like SARIMA for stock price prediction and three FinBERT infused ML models namely SVR, RFR, GBR. Eventually all predictive models are compared through regression evaluation metrics like MAE, MSE, RMSE, R2.

2025-05-01T00:00:00Z SHARDA http://dspace.dtu.ac.in:8080/jspui/handle/repository/22046 2025-08-01T06:07:52Z 2024-06-01T00:00:00Z

Title: OPTIMISING ENERGY EFFICIENCY IN WIRELESS SENSOR NETWORKS: A STUDY OF CLUSTER-BASED HIERARCHICAL ARCHITECTURE Authors: SHARDA Abstract: Wireless sensor networks (WSNs) and IoT (Internet of Things) involve numerous small sensor nodes (SNs) running on batteries, expending energy in data routing towards the sink. An efficient routing scheme is crucial for WSN longevity. In WSN routing, minimising routing hops conflicts with inter-hop routing distance - fewer hops may increase distance and vice versa. To address this, we introduce a new Multi-Objective Biogeography Based Optimization algorithm (MOBBO-R) to optimize both routing objectives and reduce energy consumption in WSNs. MOBBO-R seeks a Pareto optimal solution, prolonging WSN lifetime by enhancing routing efficiency. Validated through MATLAB, MOBBO-R outperforms past routing algorithms like PSO routing and neighbourhood routing by approximately 7% and 22%, respectively. This algorithm holds promise for various IoT and IoE-based applications globally.

2024-06-01T00:00:00Z