Abstract

Traditional methods for deep NLG adopt pipeline approaches comprising stages such as constructing syntactic input, predicting function words, linearizing the syntactic input and generating the surface forms. Though easier to visualize, pipeline approaches suffer from error propagation. In addition, information available across modules cannot be leveraged by all modules. We construct a transition-based model to jointly perform linearization, function word prediction and morphological generation, which considerably improves upon the accuracy compared to a pipelined baseline system. On a standard deep input linearization shared task, our system achieves the best results reported so far.

Abstract

Users have been trained to type keyword queries on search engines. However, recently there has been a significant rise in the number of verbose queries. Often times such queries are not well-formed. The lack of well-formedness in the query might adversely impact the downstream pipeline which processes these queries. A well-formed natural language question as a search query aids heavily in reducing errors in downstream tasks and further helps in improved query understanding. In this paper, we employ an inductive transfer learning technique by fine-tuning a pretrained language model to identify whether a search query is a well-formed natural language question or not. We show that our model trained on a recently released benchmark dataset spanning 25,100 queries gives an accuracy of 75.03% thereby improving by ∼5 absolute percentage points over the state-of-the-art.

Abstract

This paper describes our system (Fermi) for Task 5 of SemEval-2019: HatEval: Multilingual Detection of Hate Speech Against Immigrants and Women on Twitter. We participated in the subtask A for English and ranked first in the evaluation on the test set. We evaluate the quality of multiple sentence embeddings and explore multiple training models to evaluate the performance of simple yet effective embedding-ML combination algorithms. Our team - Fermi’s model achieved an accuracy of 65.00% for English language in task A. Our models, which use pretrained Universal Encoder sentence embeddings for transforming the input and SVM (with RBF kernel) for classification, scored first position (among 68) in the leaderboard on the test set for Subtask A in English language. In this paper we provide a detailed description of the approach, as well as the results obtained in the task.

Abstract

This paper describes our system (Fermi) for Task 6: OffensEval: Identifying and Categorizing Offensive Language in Social Media of SemEval-2019. We participated in all the three sub-tasks within Task 6. We evaluate multiple sentence embeddings in conjunction with various supervised machine learning algorithms and evaluate the performance of simple yet effective embedding-ML combination algorithms. Our team (Fermi)’s model achieved an F1-score of 64.40%, 62.00% and 62.60% for sub-task A, B and C respectively on the official leaderboard. Our model for subtask C which uses pretrained ELMo embeddings for transforming the input and uses SVM (RBF kernel) for training, scored third position on the official leaderboard. Through the paper we provide a detailed description of the approach, as well as the results obtained for the task.

Abstract

Online Community Question Answering Forums (cQA) have gained massive popularity within recent years. The rise in users for such forums have led to the increase in the need for automated evaluation for question comprehension and fact evaluation of the answers provided by various participants in the forum. Our team, Fermi, participated in sub-task A of Task 8 at SemEval 2019 - which tackles the first problem in the pipeline of factual evaluation in cQA forums, i.e., deciding whether a posed question asks for a factual information, an opinion/advice or is just socializing. This information is highly useful in segregating factual questions from non-factual ones which highly helps in organizing the questions into useful categories and trims down the problem space for the next task in the pipeline for fact evaluation among the available answers. Our system uses the embeddings obtained from Universal Sentence Encoder combined with XGBoost for the classification subtask A. We also evaluate other combinations of embeddings and off-the-shelf machine learning algorithms to demonstrate the efficacy of the various representations and their combinations. Our results across the evaluation test set gave an accuracy of 84% and received the first position in the final standings judged by the organizers.

Abstract

Social media platforms such as Twitter and Facebook are becoming popular in multilingual societies. This trend induces portmanteau of South Asian languages with English. The blend of multiple languages as code-mixed data has recently become popular in research communities for various NLP tasks. Code-mixed data consist of anomalies such as grammatical errors and spelling variations. In this paper, we leverage the contextual property of words where the different spelling variation of words share similar context in a large noisy social media text. We capture different variations of words belonging to same context in an unsupervised manner using distributed representations of words. Our experiments reveal that preprocessing of the code-mixed dataset based on our approach improves the performance in state-of-the-art part-of-speech tagging (POS-tagging) and sentiment analysis tasks.

Abstract

CESNA model consists of twin Bi-directional Long Short-Term Memory Recurrent Neural Networks (Bi-LSTM RNN) with shared parameters joined by a contrastive loss function based on a similarity metric. The model learns the representations of resource-poor and resource-rich language in a common emoji space by using a similarity metric based on the emojis present in sentences from both languages. The model, hence, projects sentences with similar emojis closer to each other and the sentences with different emojis farther from one another. Experiments on large-scale Twitter datasets of resource-rich languages - English and Spanish and resource-poor languages - Hindi and Telugu reveal that CESNA outperforms the state-of-the-art emoji prediction approaches based on distributional semantics,

Abstract

Bug report filing is a major part of software maintenance. Due to the asynchronous nature of the bug filing process, duplicate bug reports are filed. Detecting duplicate bug reports is an important aspect of software maintenance since the same bug should not be assigned to different developers. In this poster, we present Deep Word Embedding Network for computing similarity between two bug reports for the task of duplicate bug report detection. We propose to learn a two step model to calculate similarity between two bug reports by means of word embeddings and a deep neural network. We run experiments on two large datasets of Mozilla Project and Open Office Project and compare the proposed approach with baselines and related approaches. Through this initial work, we show that a combination of word embeddings and deep neural networks can be used to improve duplicate bug report detection.

Abstract

With news stories coming from a variety of sources, it is crucial for news aggregators to present interesting articles to the user to maximize their engagement. This creates the need for a news recommendation system which understands the content of the articles as well as accounts for the users’ preferences. Methods such as Collaborative Filtering, which are well known for general recommendations, are not suitable for news because of the short life span of articles and because of the large number of articles published each day. Apart from this, such methods do not harness the information present in the sequence in which the articles are read by the user and hence are unable to account for the specific and generic interests of the user which may keep changing with time. In order to address these issues for news recommendation, we propose the Recurrent Attentive Recommendation Engine (RARE). RARE consists of two components and utilizes the distributed representations of news articles. The first component is used to model the user’s sequential behaviour of news reading in order to understand her general interests, i.e., to get a summary of her interests. The second component utilizes an article level attention mechanism to understand her specific interests. We feed the information obtained from both the components to a Siamese Network in order to make predictions which pertain to

Abstract

Real-time Indoor theft detection from surveillancevideos is not only a challenging problem of object detection and human activity recognition in the field of computer vision, but also an urgent need for preventing theft crimes in real life. The system uses digital cameras to scan the faces of people approaching a security gate i.e. the entry gate, and then matches its faces with the faces in the database, and if found not to be a known one it automatically generates an alert. In this paper, we propose a framework for real-time indoor theft detection based on the combining result of face recognition and pattern matching by analyzing the activities with that of thieves. At last, if detected abnormal it automatically sends a message using Multimedia Message Service with the help of GPRS/GSM modem.