We develop an approach for generating deep (i.e, high-level) comprehension questions from novel text that bypasses the myriad challenges of creating a full semantic representation. We do this by decomposing the task into an ontologycrowd-relevance workflow, consisting of first representing the original text in a low-dimensional ontology, then crowdsourcing candidate question templates aligned with that space, and finally ranking potentially relevant templates for a novel region of text. If ontological labels are not available, we infer them from the text. We demonstrate the effectiveness of this method on a corpus of articles from Wikipedia alongside human judgments, and find that we can generate relevant deep questions with a precision of over 85% while maintaining a recall of 70%.

/pdf/deep-questions-without-deep-understanding-9i0pc1jxi3.pdf

Deep Questions without Deep Understanding

The Penn Discourse Treebank (PDTB) was released to the public in 2008. It remains the largest manually annotated corpus of discourse relations to date. Its focus on discourse relations that are either lexically-grounded in explicit discourse connectives or associated with sentential adjacency has not only facilitated its use in language technology and psycholinguistics but also has spawned the annotation of comparable corpora in other languages and genres.

Given this situation, this paper has four aims: (1) to provide a comprehensive introduction to the PDTB for those who are unfamiliar with it; (2) to correct some wrong (or perhaps inadvertent) assumptions about the PDTB and its annotation that may have weakened previous results or the performance of decision procedures induced from the data; (3) to explain variations seen in the annotation of comparable resources in other languages and genres, which should allow developers of future comparable resources to recognize whether the variations are relevant to them; and (4) to enumerate and explain relationships between PDTB annotation and complementary annotation of other linguistic phenomena. The paper draws on work done by ourselves and others since the corpus was released.

/pdf/reflections-on-the-penn-discourse-treebank-comparable-29u2vbr45w.pdf

Reflections on the penn discourse treebank, comparable corpora, and complementary annotation

This paper describes an automatic question generator that uses semantic pattern recognition to create questions of varying depth and type for self-study or tutoring.

/pdf/linguistic-considerations-in-automatic-question-generation-c8569vy74r.pdf

Linguistic Considerations in Automatic Question Generation

This paper is concerned with automatic generation of all possible questions from a topic of interest. Specifically, we consider that each topic is associated with a body of texts containing useful information about the topic. Then, questions are generated by exploiting the named entity information and the predicate argument structures of the sentences present in the body of texts. The importance of the generated questions is measured using Latent Dirichlet Allocation by identifying the subtopics which are closely related to the original topic in the given body of texts and applying the Extended String Subsequence Kernel to calculate their similarity with the questions. We also propose the use of syntactic tree kernels for the automatic judgment of the syntactic correctness of the questions. The questions are ranked by considering both their importance in the context of the given body of texts and syntactic correctness. To the best of our knowledge, no previous study has accomplished this task in our setting. A series of experiments demonstrate that the proposed topic-to-question generation approach can significantly outperform the state-of-the-art results.

Towards topic-to-question generation

Automatic multiple choice question (MCQ) generation from a text is a popular research area. MCQs are widely accepted for large-scale assessment in various domains and applications. However, manual generation of MCQs is expensive and time-consuming. Therefore, researchers have been attracted toward automatic MCQ generation since the late 90's. Since then, many systems have been developed for MCQ generation. We perform a systematic review of those systems. This paper presents our findings on the review. We outline a generic workflow for an automatic MCQ generation system. The workflow consists of six phases. For each of these phases, we find and discuss the list of techniques adopted in the literature. We also study the evaluation techniques for assessing the quality of the system generated MCQs. Finally, we identify the areas where the current research focus should be directed toward enriching the literature.

Automatic Multiple Choice Question Generation From Text: A Survey

In this paper, we present a system that automatically generates questions from natural language text using discourse connectives. We explore the usefulness of the discourse connectives for Question Generation (QG) that looks at the problem beyond sentence level. Our work divides the QG task into content selection and question formation. Content selection consists of finding the relevant part in text to frame question from while question formation involves sense disambiguation of the discourse connectives, identification of question type and applying syntactic transformations on the content. The system is evaluated manually for syntactic and semantic correctness.

Automatic Question Generation using Discourse Cues

Cloze questions are questions containing sentences with one or more blanks and multiple choices listed to pick an answer from. In this work, we present an automatic Cloze Question Generation (CQG) system that generates a list of important cloze questions given an English article. Our system is divided into three modules: sentence selection, keyword selection and distractor selection. We also present evaluation guidelines to evaluate CQG systems. Using these guidelines three evaluators report an average score of 3.18 (out of 4) on Cricket World Cup 2011 data.

Automatic Cloze-Questions Generation

Recent works show that the problem of noisy text normalization can be treated as a machine translation (MT) problem with convincing results. There have been supervised MT approaches which use noisy-regular parallel data for training an MT model, as well as unsupervised models which learn the translation probabilities in alternative ways and try to mimic the MT-based approach. While the supervised approaches suffer from data annotation and domain adaptation difficulties, the unsupervised models lack a holistic approach catering to all types of noise. In this paper, we propose an algorithm to artificially generate noisy text in a controlled way, from any regular English text. We see this approach as an alternative to the unsupervised approaches while getting the advantages of a parallel corpus based MT approach. We generate parallel noisy text from two widely used regular English datasets and test the MT-based approach for text normalization. Semi-supervised approaches were also tried to explore different ways of improving the parallel corpus (manually annotated) based MT approach by using the generated noisy text. An extensive analysis based on comparison of our approaches with both the supervised as well as unsupervised approaches is presented.

Experiments with artificially generated noise for cleansing noisy text

The sense of which model to use for the prediction of time series data is very important. In this paper, we have done a Project Research on Linear Regression and Long short-term memory (LSTM) model both making a prediction using the Indian Stock Market Index Nifty 50 which is a time-series data. As we know the market changes daily and daily it touches new highs and lows and huge volume change is seen as well, so can linear regression help us predict a better future prediction or LSTM helps in predicting a better future prediction is what we have tested in this project. We have given our conclusion based on the result of the RMSE. The project in a way also shows the strength of linear regression and that of are different and shows a method of implementing time series data in both the models and checking their RMSE respectively.

Linear Regression vs LSTM for Time Series Data

Heartbeat prediction is a critical area in healthcare, with early detection being essential for successful treatment. This project presents a heartbeat prediction model that utilizes Mel-Frequency Cepstral Coefficients (MFCC) and Mel Spectrogram values to classify heartbeat sounds as normal and abnormal. The Mel Spectrogram and MFCC values are extracted with the help of a python library named Librosa from the audio files of heartbeat sounds and used to train the CNN model. Convolution 2D layers are applied to train the model which is then tested on a dataset of heartbeat sounds to evaluate its performance. The model shows tremendous results in classifying heartbeat sounds, making it a valuable tool in the early detection and treatment of heart disease. The proposed model also has the potential to be incorporated into healthcare systems to automate the diagnosis and monitoring of cardiac disease in healthcare systems.

Rakshit Shah

Papers

Automatic Question Generation using Discourse Cues

Automatic Cloze-Questions Generation

Experiments with artificially generated noise for cleansing noisy text

Linear Regression vs LSTM for Time Series Data

Heartbeat Prediction using Mel Spectrogram and MFCC Value