Generative AI for Software Metadata: Overview of the Information Retrieval in Software Engineering Track at FIRE 2023

The Information Retrieval in Software Engineering (IRSE) track aims to develop solutions for automated evaluation of code comments in a machine learning framework based on human and large language model generated labels. In this track, there is a binary classification task to classify comments as useful and not useful. The dataset consists of 9048 code comments and surrounding code snippet pairs extracted from open source github C based projects and an additional dataset generated individually by teams using large language models. Overall 56 experiments have been submitted by 17 teams from various universities and software companies. The submissions have been evaluated quantitatively using the F1-Score and qualitatively based on the type of features developed, the supervised learning model used and their corresponding hyper-parameters. The labels generated from large language models increase the bias in the prediction model but lead to less over-fitted results.

Smart Knowledge Transfer using Google-like Search

To address the issue of rising software maintenance cost due to program comprehension challenges, we propose SMARTKT (Smart Knowledge Transfer), a search framework, which extracts and integrates knowledge related to various aspects of an application in form of a semantic graph. This graph supports syntax and semantic queries and converts the process of program comprehension into a {\em google-like} search problem.

Improving Contextualized Topic Models with Negative Sampling

Topic modeling has emerged as a dominant method for exploring large document collections. Recent approaches to topic modeling use large contextualized language models and variational autoencoders. In this paper, we propose a negative sampling mechanism for a contextualized topic model to improve the quality of the generated topics. In particular, during model training, we perturb the generated document-topic vector and use a triplet loss to encourage the document reconstructed from the correct document-topic vector to be similar to the input document and dissimilar to the document reconstructed from the perturbed vector. Experiments for different topic counts on three publicly available benchmark datasets show that in most cases, our approach leads to an increase in topic coherence over that of the baselines. Our model also achieves very high topic diversity.

Generation of Highlights from Research Papers Using Pointer-Generator Networks and SciBERT Embeddings

Nowadays many research articles are prefaced with research highlights to summarize the main findings of the paper. Highlights not only help researchers precisely and quickly identify the contributions of a paper, they also enhance the discoverability of the article via search engines. We aim to automatically construct research highlights given certain segments of the research paper. We use a pointer-generator network with coverage mechanism and a contextual embedding layer at the input that encodes the input tokens into SciBERT embeddings. We test our model on a benchmark dataset, CSPubSum and also present MixSub, a new multi-disciplinary corpus of papers for automatic research highlight generation. For both CSPubSum and MixSub, we have observed that the proposed model achieves the best performance compared to related variants and other models proposed in the literature. On the CSPubSum data set, our model achieves the best performance when the input is only the abstract of a paper as opposed to other segments of the paper. It produces ROUGE-1, ROUGE-2 and ROUGE-L F1-scores of 38.26, 14.26 and 35.51, respectively, METEOR F1-score of 32.62, and BERTScore F1 of 86.65 which outperform all other baselines. On the new MixSub data set, where only the abstract is the input, our proposed model (when trained on the whole training corpus without distinguishing between the subject categories) achieves ROUGE-1, ROUGE-2 and ROUGE-L F1-scores of 31.78, 9.76 and 29.3, respectively, METEOR F1-score of 24.00, and BERTScore F1 of 85.25, outperforming other models.

Melody Extraction from Polyphonic Music by Deep Learning Approaches: A Review

Melody extraction is a vital music information retrieval task among music researchers for its potential applications in education pedagogy and the music industry. Melody extraction is a notoriously challenging task due to the presence of background instruments. Also, often melodic source exhibits similar characteristics to that of the other instruments. The interfering background accompaniment with the vocals makes extracting the melody from the mixture signal much more challenging. Until recently, classical signal processing-based melody extraction methods were quite popular among melody extraction researchers. The ability of the deep learning models to model large-scale data and the ability of the models to learn automatic features by exploiting spatial and temporal dependencies inspired many researchers to adopt deep learning models for melody extraction. In this paper, an attempt has been made to review the up-to-date data-driven deep learning approaches for melody extraction from polyphonic music. The available deep models have been categorized based on the type of neural network used and the output representation they use for predicting melody. Further, the architectures of the 25 melody extraction models are briefly presented. The loss functions used to optimize the model parameters of the melody extraction models are broadly categorized into four categories and briefly describe the loss functions used by various melody extraction models. Also, the various input representations adopted by the melody extraction models and the parameter settings are deeply described. A section describing the explainability of the block-box melody extraction deep neural networks is included. The performance of 25 melody extraction methods is compared. The possible future directions to explore/improve the melody extraction methods are also presented in the paper.

Incorporating Domain Knowledge To Improve Topic Segmentation Of Long MOOC Lecture Videos

Topical Segmentation poses a great role in reducing search space of the topics taught in a lecture video specially when the video metadata lacks topic wise segmentation information. This segmentation information eases user efforts of searching, locating and browsing a topic inside a lecture video. In this work we propose an algorithm, that combines state-of-the art language model and domain knowledge graph for automatically detecting different coherent topics present inside a long lecture video. We use the language model on speech-to-text transcription to capture the implicit meaning of the whole video while the knowledge graph provides us the domain specific dependencies between different concepts of that subjects. Also leveraging the domain knowledge we can capture the way instructor binds and connects different concepts while teaching, which helps us in achieving better segmentation accuracy. We tested our approach on NPTEL lecture videos and holistic evaluation shows that it out performs the other methods described in the literature.

Knowledge Distillation for Singing Voice Detection

Singing Voice Detection (SVD) has been an active area of research in music information retrieval (MIR). Currently, two deep neural network-based methods, one based on CNN and the other on RNN, exist in literature that learn optimized features for the voice detection (VD) task and achieve state-of-the-art performance on common datasets. Both these models have a huge number of parameters (1.4M for CNN and 65.7K for RNN) and hence not suitable for deployment on devices like smartphones or embedded sensors with limited capacity in terms of memory and computation power. The most popular method to address this issue is known as knowledge distillation in deep learning literature (in addition to model compression) where a large pretrained network known as the teacher is used to train a smaller student network. However, to the best of our knowledge, such methods have not been explored yet in the domain of SVD. In this paper, efforts have been made to investigate this issue using both conventional as well as ensemble knowledge distillation techniques. Through extensive experimentation on the publicly available Jamendo dataset, we show that, not only it's possible to achieve comparable accuracies with far smaller models (upto 1000x smaller in terms of parameters), but fascinatingly, in some cases, smaller models trained with distillation, even surpass the current state-of-the-art models on voice detection performance.

Novel Radiomic Feature for Survival Prediction of Lung Cancer Patients using Low-Dose CBCT Images

Prediction of survivability in a patient for tumor progression is useful to estimate the effectiveness of a treatment protocol. In our work, we present a model to take into account the heterogeneous nature of a tumor to predict survival. The tumor heterogeneity is measured in terms of its mass by combining information regarding the radiodensity obtained in images with the gross tumor volume (GTV). We propose a novel feature called Tumor Mass within a GTV (TMG), that improves the prediction of survivability, compared to existing models which use GTV. Weekly variation in TMG of a patient is computed from the image data and also estimated from a cell survivability model. The parameters obtained from the cell survivability model are indicatives of changes in TMG over the treatment period. We use these parameters along with other patient metadata to perform survival analysis and regression. Cox's Proportional Hazard survival regression was performed using these data. Significant improvement in the average concordance index from 0.47 to 0.64 was observed when TMG is used in the model instead of GTV. The experiments show that there is a difference in the treatment response in responsive and non-responsive patients and that the proposed method can be used to predict patient survivability.

Early Response Assessment in Lung Cancer Patients using Spatio-temporal CBCT Images

We report a model to predict patient's radiological response to curative radiation therapy (RT) for non-small-cell lung cancer (NSCLC). Cone-Beam Computed Tomography images acquired weekly during the six-week course of RT were contoured with the Gross Tumor Volume (GTV) by senior radiation oncologists for 53 patients (7 images per patient). Deformable registration of the images yielded six deformation fields for each pair of consecutive images per patient. Jacobian of a field provides a measure of local expansion/contraction and is used in our model. Delineations were compared post-registration to compute unchanged ($U$), newly grown ($G$), and reduced ($R$) regions within GTV. The mean Jacobian of these regions $\mu_U$, $\mu_G$ and $\mu_R$ are statistically compared and a response assessment model is proposed. A good response is hypothesized if $\mu_R < 1.0$, $\mu_R < \mu_U$, and $\mu_G < \mu_U$. For early prediction of post-treatment response, first, three weeks' images are used. Our model predicted clinical response with a precision of $74\%$. Using reduction in CT numbers (CTN) and percentage GTV reduction as features in logistic regression, yielded an area-under-curve of 0.65 with p=0.005. Combining logistic regression model with the proposed hypothesis yielded an odds ratio of 20.0 (p=0.0).

Posture and sequence recognition for Bharatanatyam dance performances using machine learning approach

Understanding the underlying semantics of performing arts like dance is a challenging task. Dance is multimedia in nature and spans over time as well as space. Capturing and analyzing the multimedia content of the dance is useful for the preservation of cultural heritage, to build video recommendation systems, to assist learners to use tutoring systems. To develop an application for dance, three aspects of dance analysis need to be addressed: 1) Segmentation of the dance video to find the representative action elements, 2) Matching or recognition of the detected action elements, and 3) Recognition of the dance sequences formed by combining a number of action elements under certain rules. This paper attempts to solve three fundamental problems of dance analysis for understanding the underlying semantics of dance forms. Our focus is on an Indian Classical Dance (ICD) form known as Bharatanatyam. As dance is driven by music, we use the music as well as motion information for key posture extraction. Next, we recognize the key postures using machine learning as well as deep learning techniques. Finally, the dance sequence is recognized using the Hidden Markov Model (HMM). We capture the multi-modal data of Bharatanatyam dance using Kinect and build an annotated data set for research in ICD.