Browsing by Author "Bansal, Ritesh"

Now showing 1 - 2 of 2
  • Thumbnail Image
    Big Data Text Summarization: Using Deep Learning to Summarize Theses and Dissertations
    Ahuja, Naman; Bansal, Ritesh; Ingram, William A.; Jude, Palakh; Kahu, Sampanna; Wang, Xinyue (Virginia Tech, 2018-12-05)
    Team 16 in the fall 2018 course "CS 4984/5984 Big Data Text Summarization," in partnership with the University Libraries and the Digital Library Research Laboratory, prepared a corpus of electronic theses and dissertations (ETDs) for students to study natural language processing with the power of state-of-the-art deep learning technology. The ETD corpus is made up of 13,071 doctoral dissertations and 17,890 master theses downloaded from the University Libraries’ VTechWorks system. This particular study is designed to explore big data summarization for ETDs, which is a relatively under-explored area. The result of the project will help to address the difficulty of information extraction from ETD documents, the potential of transfer learning on automatic summarization of ETD chapters, and the quality of state-of-the-art deep learning summarization technologies when applied to the ETD corpus. The goal of this project is to generate chapter level abstractive summaries for an ETD collection through deep learning. Major challenges of the project include accurately extracting well-formatted chapter text from PDF files, and the lack of labeled data for supervised deep learning models. For PDF processing, we compare two state of the art scholarly PDF data extraction tools, Grobid and Science-Parse, which generate structured documents from which we can further extract metadata and chapter level text. For the second challenge, we perform transfer learning by training supervised learning models on a labeled dataset of Wikipedia articles related to the ETD collection. Our experimental models include Sequence-to-Sequence and Pointer Generator summarization models. Besides supervised models, we also experiment with an unsupervised reinforcement model, Fast Abstractive Summarization-RL. The general pipeline for our experiments consists of the following steps: PDF data processing and chapter extraction, collecting a training data set of Wikipedia articles, manually creating human generated gold standard summaries for testing and validation, building deep learning models for chapter summarization, evaluating and tuning the models based on results, and then iteratively refining the whole process.
  • Thumbnail Image
    Toward an Intelligent Crawling Scheduler for Archiving News Websites Using Reinforcement Learning
    Wang, Xinyue; Ahuja, Naman; Llorens, Nathaniel; Bansal, Ritesh; Dhar, Siddharth (Virginia Tech, 2019-12-03)
    Web crawling is one of the fundamental activities for many kinds of web technology organizations and companies such as Internet Archive and Google. While companies like Google often focus on content delivery for users, web archiving organizations such as the Internet Archive pay more attention to the accurate preservation of the web. Crawling accuracy and efficiency are major concerns in this task. An ideal crawling module should be able to keep up with the changes in the target web site with minimal crawling frequency to maximize the routine crawling efficiency. In this project, we investigate using information from web archives' history to help the crawling process within the scope of news websites. We aim to build a smart crawling module that can predict web content change accurately both on the web page and web site structure level through modern machine learning algorithms and deep learning architectures. At the end of the project: We have collected and processed raw web archive collections from Archive.org and through our frequent crawling jobs. We have developed methods to extract identical copies of web page content and web site structure from the web archive data. We have implemented baseline models for predicting web page content change and web site structure change, web page content change with supervised machine learning algorithms; We have implemented two different reinforcement learning models for generating a web page crawling plan: a continuous prediction model and a sparse prediction model. Our results show that the reinforcement learning modal has the potential to work as an intelligent web crawling scheduler.