Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Personalized Summarization of Scientific Scholarly Texts

Alka Khurana, Vasudha Bhatnagar, Vikas Kumar

TL;DR

P-Summ algorithm, an unsupervised algorithm that generates an extractive summary of scientific scholarly text to meet the personal knowledge needs of the user, is presented and has the capability to meet negative (or positive) knowledge preferences of the users.

Abstract

In this paper, we present a proposal for an unsupervised algorithm, P-Summ, that generates an extractive summary of scientific scholarly text to meet the personal knowledge needs of the user. The method delves into the latent semantic space of the document exposed by Weighted Non-negative Matrix Factorization, and scores sentences in consonance with the knowledge needs of the user. The novelty of the algorithm lies in its ability to include desired knowledge and eliminate unwanted knowledge in the personal summary. We also propose a multi-granular evaluation framework, which assesses the quality of generated personal summaries at three levels of granularity - sentence, terms and semantic. The framework uses system generated generic summary instead of human generated summary as gold standard for evaluating the quality of personal summary generated by the algorithm. The effectiveness of the algorithm at the semantic level is evaluated by taking into account the reference summary and the knowledge signals. We evaluate the performance of P-Summ algorithm over four data-sets consisting of scientific articles. Our empirical investigations reveal that the proposed method has the capability to meet negative (or positive) knowledge preferences of the user.

Personalized Summarization of Scientific Scholarly Texts

TL;DR

P-Summ algorithm, an unsupervised algorithm that generates an extractive summary of scientific scholarly text to meet the personal knowledge needs of the user, is presented and has the capability to meet negative (or positive) knowledge preferences of the users.

Abstract

In this paper, we present a proposal for an unsupervised algorithm, P-Summ, that generates an extractive summary of scientific scholarly text to meet the personal knowledge needs of the user. The method delves into the latent semantic space of the document exposed by Weighted Non-negative Matrix Factorization, and scores sentences in consonance with the knowledge needs of the user. The novelty of the algorithm lies in its ability to include desired knowledge and eliminate unwanted knowledge in the personal summary. We also propose a multi-granular evaluation framework, which assesses the quality of generated personal summaries at three levels of granularity - sentence, terms and semantic. The framework uses system generated generic summary instead of human generated summary as gold standard for evaluating the quality of personal summary generated by the algorithm. The effectiveness of the algorithm at the semantic level is evaluated by taking into account the reference summary and the knowledge signals. We evaluate the performance of P-Summ algorithm over four data-sets consisting of scientific articles. Our empirical investigations reveal that the proposed method has the capability to meet negative (or positive) knowledge preferences of the user.
Paper Structure (31 sections, 3 equations, 9 figures, 6 tables, 2 algorithms)

This paper contains 31 sections, 3 equations, 9 figures, 6 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Personalized Summarization: Challenges
  3. Contributions
  4. Personalized Summarization: The Problem and Approaches
  5. Related Works
  6. Based on mode of Knowledge Signals
  7. Based on Approach for Summarization
  8. Personalized Summarization - Handling Positive and Negative signals
  9. Weighted Non-negative Matrix Factorization
  10. Penalty and Reward for Knowledge Signals
  11. Identify Semantically related Terms
  12. Setting the Weight Matrix
  13. Number of latent factors
  14. P-Summ – The Proposed Algorithm
  15. Evaluation Framework for Personal Summaries
  16. ...and 16 more sections

Figures (9)

  • Figure 1: Inputs to P-Summ algorithm and its core steps
  • Figure 2: Approaches for Personalized Summarization
  • Figure 3: Approach for computing weight matrix $W$
  • Figure 4: Density plot for semantic similarities between $\mathcal{K^-}$ and, personal and generic summaries for AIPubSumm dataset for 25% summary length and one keyword.
  • Figure 5: Density plot for semantic similarities between $\mathcal{K^+}$ and, personal and generic summaries for AIPubSumm dataset
  • ...and 4 more figures

Theorems & Definitions (1)

  • Example 1