Sentiment-Aware Extractive and Abstractive Summarization for Unstructured Text Mining
2025
Read Sentiment-Aware Extractive and Abstractive Summarization for Unstructured Text Mining (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-20404.txt.
Introduction
Junyi Liu and Stanley Kok summarize messy, emotional text while keeping the feelings in the summary. Social media posts, reviews, and forum threads are short, noisy, and fragmented, unlike the tidy news articles most summarizers were built for. Yet for brand monitoring and market analysis, the sentiment is often the whole point.
Their approach adds sentiment signals to both an extractive method, TextRank, and an abstractive one, UniLM. Sentiment shapes which sentences get picked and how new summary text is generated. The result is concise summaries that preserve emotional nuance and thematic relevance. For information systems work, such summaries promise faster interventions and sharper strategic decisions in fast-moving online environments.
Catalog record
| YEAR | 2025 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 21 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2512-20404
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Sentiment-Aware Extractive and Abstractive Summarization for Unstructured Text Mining Abstract \researchtype Regular Paper Sentiment-Aware Summarization for Unstructured Text \conference Workshop on Information Technologies and Systems, Nashville, USA 2025 Junyi Liu Stanley Kok Department of Information Systems and Analytics Department of Information Systems and Analytics National University of Singapore National University of Singapore [email protected] [email protected] 1 Introduction With the growing volume of unstructured textual data—such as social media posts, online reviews, forum discussions, and news articles—text mining has become increasingly crucial in Information Systems ( 5 ) . It enables the automated extraction of implicit, previously unknown, and potentially valuable knowledge from vast datasets of unstructured data. Text mining techniques, including sentiment analysis, topic modeling, and text classification, equip organizations with the scalability and repeatability needed to process large datasets efficiently. Consequently, text mining has become an invaluable resource for businesses to understand consumer preferences, identify emerging trends, and respond to public sentiment in real-time. Information systems (IS) research has increasingly leveraged text mining for organizational insight. For example, 14 examined how incentivized reviews affect textual coherence; 2 introduced sentence-level topic modeling; and 3 incorporated sentiment signals into topic inference, highlighting how emotional undercurrents shape user opinions and inform business strategy. However, a significant challenge is handling the vast and fragmented unstructured data from platforms like Weibo, Instagram, and Reddit. The discussions on these platforms are wide-ranging, fragmented, and loosely structured, making it difficult for traditional text mining methods to efficiently extract useful information, especially in terms of capturing emotions and thematic contexts. Text summarization offers a natural solution for handling vast and fragmented text by condensing lengthy content into concise summaries. It generally involves two approaches: extractive, which selects important sentences directly from the text, and abstractive, which generates new sentences that capture the essence of the original content. Existing summarization methods are optimized for structured texts like news articles but struggle with informal social media content. Social media often contains fragmented text and emotional expressions, making summarization more challenging. In Information Systems, understanding emotional cues is crucial for brand monitoring, customer interaction, and market analysis. Emotionally charged comments can reveal consumer feedback and potential issues, while ignoring these signals may lead to missed insights. Although sentiment analysis has gained traction in IS for detecting emotional polarity and intensity ( 7 ; 13 ) , few studies have integrated sentiment modeling with summarization in the context of short, noisy, user posts. In practice, decision-makers need to understand not only what users discuss, but how they feel—whether satisfied, frustrated, or indifferent. Sentiment-enriched summaries can thus provide actionable insights, enabling timely interventions and strategic responses. To address these challenges, we propose a sentiment-aware summarization framework for unstructured text mining. We extend both extractive summarization (e.g., graph-based TextRank) and abstractive summarization (e.g., Transformer-based UniLM) by integrating sentiment signals into their core mechanisms. This dual approach provides feasible solutions for improving both extractive and abstractive summarization methods, offering valuable contributions to text mining and Information Systems, particularly in enhancing the ability to capture emotional nuances and thematic relevance in user-generated content for decision support and strategic insights. 2 Related Work 2.1 Text Mining in IS Text mining plays a crucial role in Information Systems (IS) research, especially due to its ability to analyze large volumes of unstructured textual data and extract valuable insights. Studies have investigated how text mining can help organizations understand consumer sentiment, inform business strategies, and improve predictive modeling. For example, research has explored how sentiment analysis of online reviews can influence purchasing behavior ( 18 ) , how topic modeling on business pages can predict engagement patterns ( 17 ) , and how text classification methods applied to social media data can forecast trends more accurately than traditional approaches ( 16 ) . Additionally, studies have examined how text mining techniques can track shifts in customer opinions ( 8 ) , assess competition within digital content ecosystems ( 15 ) , and measure the impact of external incentives on user-generated content quality ( 11 ) . While these studies highlight the strategic potential of text mining, many focus on analyzing data volume, user behavior, or predictive accuracy. However, there has been less emphasis on developing automated methods for synthesizing short-form, fragmented, and sentiment-rich text—despite its increasing presence on platforms such as Reddit, Twitter, or Weibo. This gap limits the full potential of text mining in decision support systems (DSS), where decision-makers often need to interpret large datasets of user opinions quickly and accurately. 2.2 Sentiment Analysis in IS Sentiment analysis has emerged as a powerful tool for understanding emotional cues in textual data. Within IS, prior work has addressed domain adaptation for sentiment classification ( 7 ) and examined how emotional tone influences content consumption and sharing behavior ( 13 ) . For instance, negative sentiment has been shown to attract more attention on news websites, while positive sentiment increases shareability on social platforms. Despite these advances, sentiment analysis is rarely integrated with summarization in IS research. Most approaches treat sentiment classification and summarization as separate tasks, overlooking their joint potential to distill both the content and emotional tone of UGC. This disconnect is particularly limiting in settings where the emotional valence of content—such as frustration, enthusiasm, or sarcasm—is key to interpreting user concerns or opportunities. 3 Background 3.1 Text Summarization Text summarization aims to distill the essential content of a document into a concise version. Existing methods are broadly categorized into extractive and abstractive approaches. Extractive methods identify and select salient sentences from the input, ensuring grammaticality but often including redundancy. Abstractive methods generate new sentences that capture the core meaning, offering greater flexibility but requiring robust language modeling capabilities. Hybrid approaches aim to balance the strengths of both. 3.2 Textrank TextRank is a graph-based algorithm widely used for extractive summarization ( 12 ) . It treats the sentences of a document as nodes in a graph, where edges represent the similarity between sentences. The importance of each sentence is determined by its centrality within the graph. Formally, given a document D=\{s_{1},s_{2},\dots,s_{N}\} , a graph G=(V,E) is constructed, where V represents the sentences in D and E represents the edges connecting similar sentences. The weight of an edge between two sentences s_{i} and s_{j} , denoted as w_{ij} , is calculated as: w_{ij}=\frac{\text{sim}(s_{i},s_{j})}{\sum_{k\neq i}\text{sim}(s_{i},s_{k})}, where \text{sim}(s_{i},s_{j}) is a similarity function (e.g., cosine similarity or Jaccard index) that measures the relationship between sentences. The importance score p_{i} of a sentence s_{i} is computed iteratively: p_{i}=(1-d)+d\cdot\sum_{j\in N(i)}\frac{w_{ij}\cdot p_{j}}{\sum_{k\in N(j)}w_{jk}}, where d is a damping factor (typically set to 0.85), and N(i) represents the neighboring sentences of s_{i} in the graph. Sentences with the highest importance scores are selected for the summary. 3.3 UniLM UniLM (Unified Language Model) is a pre-trained transformer model designed for both understanding and generation tasks ( 6 ) . In the context of summarization, UniLM generates abstractive summaries by conditioning on the input document. UniLM adopts a unified language modeling framework, in which different self-attention masks are used during training to support various language modeling objectives. For abstractive summarization, UniLM is fine-tuned in an autoregressive sequence-to-sequence setting. Given an input document D and a summary sequence S=(s_{1},s_{2},\dots,s_{|S|}) , the model is trained to maximize the conditional likelihood