Mukesh Mohania

Dikshant Kukreja, Kshitij Sah, Karan Goyal et al.

Educational diagrams -- labeled illustrations of biological processes, chemical structures, physical systems, and mathematical concepts -- are essential cognitive tools in K-12 instruction. Yet no existing method can generate them both accurately and engagingly. Open-source diffusion models produce visually rich images but catastrophically garble text labels. Code-based generation via LLMs guarantees label correctness but yields visually flat outputs. Closed-source APIs partially bridge this gap but remain unreliable and prohibitively expensive at educational scale. We quantify this accuracy-aesthetics dilemma across all three paradigms on 400 K-12 diagram prompts, measuring both label fidelity and visual quality through complementary automated and human evaluation protocols. To resolve it, we propose CAGE (Code-Anchored Generative Enhancement): an LLM synthesizes executable code producing a structurally correct diagram, then a diffusion model conditioned on the programmatic output via ControlNet refines it into a visually polished graphic while preserving label fidelity. We also introduce EduDiagram-2K, a collection of 2,000 paired programmatic-stylized diagrams enabling this pipeline, and present proof-of-concept results and a research agenda for the multimedia community.

Vasu Goel, Dhruv Sahnan, Venktesh V et al.

Online education platforms are powered by various NLP pipelines, which utilize models like BERT to aid in content curation. Since the inception of the pre-trained language models like BERT, there have also been many efforts toward adapting these pre-trained models to specific domains. However, there has not been a model specifically adapted for the education domain (particularly K-12) across subjects to the best of our knowledge. In this work, we propose to train a language model on a corpus of data curated by us across multiple subjects from various sources for K-12 education. We also evaluate our model, K12-BERT, on downstream tasks like hierarchical taxonomy tagging.

Dikshant Kukreja, Kshitij Sah, Karan Goyal et al.

When asked to describe a molecular diagram, a Vision-Language Model correctly identifies ``a benzene ring with an -OH group.'' When asked to reason about the same image, it answers incorrectly. The model can see but it cannot think about what it sees. We term this the perception-integration gap: a failure where visual information is successfully extracted but lost during downstream reasoning, invisible to single-configuration benchmarks that conflate perception with integration under one accuracy number. To systematically expose such failures, we introduce DISSECT, a 12,000-question diagnostic benchmark spanning Chemistry (7,000) and Biology (5,000). Every question is evaluated under five input modes -- Vision+Text, Text-Only, Vision-Only, Human Oracle, and a novel Model Oracle in which the VLM first verbalizes the image and then reasons from its own description -- yielding diagnostic gaps that decompose performance into language-prior exploitation, visual extraction, perception fidelity, and integration effectiveness. Evaluating 18~VLMs, we find that: (1) Chemistry exhibits substantially lower language-prior exploitability than Biology, confirming molecular visual content as a harder test of genuine visual reasoning; (2) Open-source models consistently score higher when reasoning from their own verbalized descriptions than from raw images, exposing a systematic integration bottleneck; and (3) Closed-source models show no such gap, indicating that bridging perception and integration is the frontier separating open-source from closed-source multimodal capability. The Model Oracle protocol is both model and benchmark agnostic, applicable post-hoc to any VLM evaluation to diagnose integration failures.

Venktesh V, Md. Shad Akhtar, Mukesh Mohania et al.

Online learning platforms conduct exams to evaluate the learners in a monotonous way, where the questions in the database may be classified into Bloom's Taxonomy as varying levels in complexity from basic knowledge to advanced evaluation. The questions asked in these exams to all learners are very much static. It becomes important to ask new questions with different difficulty levels to each learner to provide a personalized learning experience. In this paper, we propose a multi-task method with an interactive attention mechanism, Qdiff, for jointly predicting Bloom's Taxonomy and difficulty levels of academic questions. We model the interaction between the predicted bloom taxonomy representations and the input representations using an attention mechanism to aid in difficulty prediction. The proposed learning method would help learn representations that capture the relationship between Bloom's taxonomy and difficulty labels. The proposed multi-task method learns a good input representation by leveraging the relationship between the related tasks and can be used in similar settings where the tasks are related. The results demonstrate that the proposed method performs better than training only on difficulty prediction. However, Bloom's labels may not always be given for some datasets. Hence we soft label another dataset with a model fine-tuned to predict Bloom's labels to demonstrate the applicability of our method to datasets with only difficulty labels.

Venktesh Viswanathan, Mukesh Mohania, Vikram Goyal

Online learning systems have multiple data repositories in the form of transcripts, books and questions. To enable ease of access, such systems organize the content according to a well defined taxonomy of hierarchical nature (subject-chapter-topic). The task of categorizing inputs to the hierarchical labels is usually cast as a flat multi-class classification problem. Such approaches ignore the semantic relatedness between the terms in the input and the tokens in the hierarchical labels. Alternate approaches also suffer from class imbalance when they only consider leaf level nodes as labels. To tackle the issues, we formulate the task as a dense retrieval problem to retrieve the appropriate hierarchical labels for each content. In this paper, we deal with categorizing questions. We model the hierarchical labels as a composition of their tokens and use an efficient cross-attention mechanism to fuse the information with the term representations of the content. We also propose an adaptive in-batch hard negative sampling approach which samples better negatives as the training progresses. We demonstrate that the proposed approach \textit{TagRec++} outperforms existing state-of-the-art approaches on question datasets as measured by Recall@k. In addition, we demonstrate zero-shot capabilities of \textit{TagRec++} and ability to adapt to label changes.

Rishabh Gupta, Venktesh V, Mukesh Mohania et al.

This paper introduces the novel task of scoring paraphrases for Algebraic Word Problems (AWP) and presents a self-supervised method for doing so. In the current online pedagogical setting, paraphrasing these problems is helpful for academicians to generate multiple syntactically diverse questions for assessments. It also helps induce variation to ensure that the student has understood the problem instead of just memorizing it or using unfair means to solve it. The current state-of-the-art paraphrase generation models often cannot effectively paraphrase word problems, losing a critical piece of information (such as numbers or units) which renders the question unsolvable. There is a need for paraphrase scoring methods in the context of AWP to enable the training of good paraphrasers. Thus, we propose ParaQD, a self-supervised paraphrase quality detection method using novel data augmentations that can learn latent representations to separate a high-quality paraphrase of an algebraic question from a poor one by a wide margin. Through extensive experimentation, we demonstrate that our method outperforms existing state-of-the-art self-supervised methods by up to 32% while also demonstrating impressive zero-shot performance.

Rishabh Gupta, Venktesh V., Mukesh Mohania et al.

In this paper, we propose SCANING, an unsupervised framework for paraphrasing via controlled noise injection. We focus on the novel task of paraphrasing algebraic word problems having practical applications in online pedagogy as a means to reduce plagiarism as well as ensure understanding on the part of the student instead of rote memorization. This task is more complex than paraphrasing general-domain corpora due to the difficulty in preserving critical information for solution consistency of the paraphrased word problem, managing the increased length of the text and ensuring diversity in the generated paraphrase. Existing approaches fail to demonstrate adequate performance on at least one, if not all, of these facets, necessitating the need for a more comprehensive solution. To this end, we model the noising search space as a composition of contextual and syntactic aspects and sample noising functions consisting of either one or both aspects. This allows for learning a denoising function that operates over both aspects and produces semantically equivalent and syntactically diverse outputs through grounded noise injection. The denoising function serves as a foundation for learning a paraphrasing function which operates solely in the input-paraphrase space without carrying any direct dependency on noise. We demonstrate SCANING considerably improves performance in terms of both semantic preservation and producing diverse paraphrases through extensive automated and manual evaluation across 4 datasets.

Maksimjeet Chowdhary, Sanyam Goyal, Venktesh V et al.

Online learning platforms provide diverse questions to gauge the learners' understanding of different concepts. The repository of questions has to be constantly updated to ensure a diverse pool of questions to conduct assessments for learners. However, it is impossible for the academician to manually skim through the large repository of questions to check for duplicates when onboarding new questions from external sources. Hence, we propose a tool QDup in this paper that can surface near-duplicate and semantically related questions without any supervised data. The proposed tool follows an unsupervised hybrid pipeline of statistical and neural approaches for incorporating different nuances in similarity for the task of question duplicate detection. We demonstrate that QDup can detect near-duplicate questions and also suggest related questions for practice with remarkable accuracy and speed from a large repository of questions. The demo video of the tool can be found at https://www.youtube.com/watch?v=loh0_-7XLW4.

Aarish Chhabra, Nandini Bansal, Venktesh V et al.

Exams are conducted to test the learner's understanding of the subject. To prevent the learners from guessing or exchanging solutions, the mode of tests administered must have sufficient subjective questions that can gauge whether the learner has understood the concept by mandating a detailed answer. Hence, in this paper, we propose a novel hybrid unsupervised approach leveraging rule-based methods and pre-trained dense retrievers for the novel task of automatically converting the objective questions to subjective questions. We observe that our approach outperforms the existing data-driven approaches by 36.45% as measured by Recall@k and Precision@k.

Karan Goyal, Dikshant Kukreja, Vikram Goyal et al.

Proper citation of relevant literature is essential for contextualising and validating scientific contributions. While current citation recommendation systems leverage local and global textual information, they often overlook the nuances of the human citation behaviour. Recent methods that incorporate such patterns improve performance but incur high computational costs and introduce systematic biases into downstream rerankers. To address this, we propose Profiler, a lightweight, non-learnable module that captures human citation patterns efficiently and without bias, significantly enhancing candidate retrieval. Furthermore, we identify a critical limitation in current evaluation protocol: the systems are assessed in a transductive setting, which fails to reflect real-world scenarios. We introduce a rigorous Inductive evaluation setting that enforces strict temporal constraints, simulating the recommendation of citations for newly authored papers in the wild. Finally, we present DAVINCI, a novel reranking model that integrates profiler-derived confidence priors with semantic information via an adaptive vector-gating mechanism. Our system achieves new state-of-the-art results across multiple benchmark datasets, demonstrating superior efficiency and generalisability.

Venktesh V, Mukesh Mohania, Vikram Goyal

Keyphrase extraction from a given document is the task of automatically extracting salient phrases that best describe the document. This paper proposes a novel unsupervised graph-based ranking method to extract high-quality phrases from a given document. We obtain the contextualized embeddings from pre-trained language models enriched with topic vectors from Latent Dirichlet Allocation (LDA) to represent the candidate phrases and the document. We introduce a scoring mechanism for the phrases using the information obtained from contextualized embeddings and the topic vectors. The salient phrases are extracted using a ranking algorithm on an undirected graph constructed for the given document. In the undirected graph, the nodes represent the phrases, and the edges between the phrases represent the semantic relatedness between them, weighted by a score obtained from the scoring mechanism. To demonstrate the efficacy of our proposed method, we perform several experiments on open source datasets in the science domain and observe that our novel method outperforms existing unsupervised embedding based keyphrase extraction methods. For instance, on the SemEval2017 dataset, our method advances the F1 score from 0.2195 (EmbedRank) to 0.2819 at the top 10 extracted keyphrases. Several variants of the proposed algorithm are investigated to determine their effect on the quality of keyphrases. We further demonstrate the ability of our proposed method to collect additional high-quality keyphrases that are not present in the document from external knowledge bases like Wikipedia for enriching the document with newly discovered keyphrases. We evaluate this step on a collection of annotated documents. The F1-score at the top 10 expanded keyphrases is 0.60, indicating that our algorithm can also be used for 'concept' expansion using external knowledge.

Kshitij Gulati, Gaurav Verma, Mukesh Mohania et al.

With the global metamorphosis of the beauty industry and the rising demand for beauty products worldwide, the need for an efficacious makeup recommendation system has never been more. Despite the significant advancements made towards personalised makeup recommendation, the current research still falls short of incorporating the context of occasion in makeup recommendation and integrating feedback for users. In this work, we propose BeautifAI, a novel makeup recommendation system, delivering personalised occasion-oriented makeup recommendations to users while providing real-time previews and continuous feedback. The proposed work's novel contributions, including the incorporation of occasion context, region-wise makeup recommendation, real-time makeup previews and continuous makeup feedback, set our system apart from the current work in makeup recommendation. We also demonstrate our proposed system's efficacy in providing personalised makeup recommendation by conducting a user study.

Venktesh V, Mukesh Mohania, Vikram Goyal

Online educational platforms organize academic questions based on a hierarchical learning taxonomy (subject-chapter-topic). Automatically tagging new questions with existing taxonomy will help organize these questions into different classes of hierarchical taxonomy so that they can be searched based on the facets like chapter. This task can be formulated as a flat multi-class classification problem. Usually, flat classification based methods ignore the semantic relatedness between the terms in the hierarchical taxonomy and the questions. Some traditional methods also suffer from the class imbalance issues as they consider only the leaf nodes ignoring the hierarchy. Hence, we formulate the problem as a similarity-based retrieval task where we optimize the semantic relatedness between the taxonomy and the questions. We demonstrate that our method helps to handle the unseen labels and hence can be used for taxonomy tagging in the wild. In this method, we augment the question with its corresponding answer to capture more semantic information and then align the question-answer pair's contextualized embedding with the corresponding label (taxonomy) vector representations. The representations are aligned by fine-tuning a transformer based model with a loss function that is a combination of the cosine similarity and hinge rank loss. The loss function maximizes the similarity between the question-answer pair and the correct label representations and minimizes the similarity to unrelated labels. Finally, we perform experiments on two real-world datasets. We show that the proposed learning method outperforms representations learned using the multi-class classification method and other state of the art methods by 6% as measured by Recall@k. We also demonstrate the performance of the proposed method on unseen but related learning content like the learning objectives without re-training the network.

Nebula Alam, Khoi-Nguyen Tran, Sue Ann Chen et al.

The importance of patents is well recognised across many regions of the world. Many patent mining systems have been proposed, but with limited predictive capabilities. In this demo, we showcase how predictive algorithms leveraging the state-of-the-art machine learning and deep learning techniques can be used to improve understanding of patents for inventors, patent evaluators, and business analysts alike. Our demo video is available at http://ibm.biz/ecml2019-demo-patent-analytics

Khoi-Nguyen Tran, Jey Han Lau, Danish Contractor et al.

Instructional Systems Design is the practice of creating of instructional experiences that make the acquisition of knowledge and skill more efficient, effective, and appealing. Specifically in designing courses, an hour of training material can require between 30 to 500 hours of effort in sourcing and organizing reference data for use in just the preparation of course material. In this paper, we present the first system of its kind that helps reduce the effort associated with sourcing reference material and course creation. We present algorithms for document chunking and automatic generation of learning objectives from content, creating descriptive content metadata to improve content-discoverability. Unlike existing methods, the learning objectives generated by our system incorporate pedagogically motivated Bloom's verbs. We demonstrate the usefulness of our methods using real world data from the banking industry and through a live deployment at a large pharmaceutical company.