The work presented here describes our participation in DISRPT 2025 shared task in three tasks, Task1: Discourse Unit Segmentation across Formalisms, Task 2: Discourse Connective Identification across Languages and Task 3: Discourse Relation Classification across Formalisms. We have fine-tuned XLM-RoBERTa, a language model to address these three tasks. We have come up with one single multilingual language model for each task. Our system handles data in both the formats .conllu and .tok and different discourse formalisms. We have obtained encouraging results. The performance on test data in the three tasks is similar to the results obtained for the development data.

This paper describes an approach on an end to end model for Multilingual Coreference Resolution (CR) for low resource languages such as Tamil, Malayalam and Hindi. We have done fine tune the XLM-Roberta large model on multilingual training dataset using specific languages with linguistic features and without linguistic features. XLM-R with linguistic features achieves better results than the baseline system. This shows that giving the linguistic knowledge enriches the system performance. The performance of the system is comparable with the state of the art systems.

End-to-end coreference resolution is the task of identifying the mentions in a text that refer to the same real world entity and grouping them into clusters. It is crucially required for natural language understanding tasks and other high-level NLP tasks. In this paper, we present an end-to-end architecture for neural coreference resolution using AdapterFusion, a new two stage learning algorithm that leverages knowledge from multiple tasks. First task is in identifying the mentions in the text and the second to determine the coreference clusters. In the first task we learn task specific parameters called adapters that encapsulate the taskspecific information and then combine the adapters in a separate knowledge composition step to identify the mentions and their clusters. We evaluated it using FIRE corpus for Malayalam and Tamil and we achieved state of art performance.

Neural machine translation (NMT) has achieved state-of-art performance in high-resource language pairs, but the performance of NMT drops in low-resource conditions. Morphologically rich languages are yet another challenge in NMT. The common strategy to handle this issue is to apply sub-word segmentation. In this work, we compare the morphologically inspired segmentation methods against the Byte Pair Encoding (BPE) in processing the input for building NMT systems for Hindi to Malayalam and Hindi to Tamil, where Hindi is an Indo-Aryan language and Malayalam and Tamil are south Dravidian languages. These two languages are low resource, morphologically rich and agglutinative. Malayalam is more agglutinative than Tamil. We show that for both the language pairs, the morphological segmentation algorithm out-performs BPE. We also present an elaborate analysis on translation outputs from both the NMT systems.

Dependency parsing is the process of analysing the grammatical structure of a sentence based on the dependencies between the words in a sentence. The annotation of dependency parsing is done using different formalisms at word-level namely Universal Dependencies and chunk-level namely AnnaCorra. Though dependency parsing is deeply dealt in languages such as English, Czech etc the same cannot be adopted for the morphologically rich and agglutinative languages. In this paper, we discuss the development of a dependency parser for Tamil, a South Dravidian language. The different characteristics of the language make this task a challenging task. Tamil, a morphologically rich and agglutinative language, has copula drop, accusative and genitive case drop and pro-drop. Coordinative constructions are introduced by affixation of morpheme ‘um’. Embedded clausal structures are common in relative participle and complementizer clauses. In this paper, we have discussed our approach to handle some of these challenges. We have used Malt parser, a supervised learning- approach based implementation. We have obtained an accuracy of 79.27% for Unlabelled Attachment Score, 73.64% for Labelled Attachment Score and 68.82% for Labelled Accuracy.

Natural language understanding by automatic tools is the vital requirement for document processing tools. To achieve it, automatic system has to understand the coherence in the text. Co-reference chains bring coherence to the text. The commonly occurring reference markers which bring cohesiveness are Pronominal, Reflexives, Reciprocals, Distributives, One-anaphors, Noun–noun reference. Here in this paper, we deal with noun-noun reference in Tamil. We present the methodology to resolve these noun-noun anaphors and also present the challenges in handling the noun-noun anaphoric relations in Tamil.