LaMSUM: Amplifying Voices Against Harassment through LLM Guided Extractive Summarization of User Incident Reports

# LaMSUM: Amplifying Voices Against Harassment through LLM Guided Extractive Summarization of User Incident Reports

Source: https://arxiv.org/html/2406.15809

###### Abstract

Citizen reporting platforms help the public and authorities stay informed about sexual harassment incidents. However, the high volume of data shared on these platforms makes reviewing each individual case challenging. Therefore, a summarization algorithm capable of processing and understanding various code-mixed languages is essential. In recent years, Large Language Models (LLMs) have shown exceptional performance in NLP tasks, including summarization. LLMs inherently produce abstractive summaries by paraphrasing the original text, while the generation of extractive summaries – selecting specific subsets from the original text – through LLMs remains largely unexplored. Moreover, LLMs have a limited context window size, restricting the amount of data that can be processed at once. We tackle these challenges by introducing LaMSUM, a novel multi-level framework combining summarization with different voting methods to generate extractive summaries for large collections of incident reports using LLMs. Extensive evaluation using four popular LLMs (Llama, Mistral, Claude and GPT-4o) demonstrates that LaMSUM outperforms state-of-the-art extractive summarization methods. Overall, this work represents one of the first attempts to achieve extractive summarization through LLMs, and is likely to support stakeholders by offering a comprehensive overview and enabling them to develop effective policies to minimize incidents of unwarranted harassment.

**Warning:** This paper contains content that may be disturbing or upsetting.

## 1 Introduction

| Category | Post |
|----------|------|
| Robbery | This incident took place in the evening. Two bikers came on a bike and snatched Rs. 17000 from an old lady at gun point. |
| Stalking | I was stalked by a guy who followed me for days and also he sent me letters on my doorstep saying he was madly in love with me and he is obsessed with my body. |
| Sexual Invites | I was walking on footpath to a place nearby to meet my friends. A man was driving his car on parallel road and was continuously passing vulgar comments. I ignored but 15 min later he stopped his car and said "chalri h ky, paise h mere pas" (I have money, you want to come with me?). I screamed and asked for help from people around me. |
| Masturbation in public | One day I reached my school prior to the school timings mistakenly. One of the drivers called me and my friend and started masturbating in front of us. |
| Ogling | I was going to my coaching by driving my scooty when suddenly some boys came on a bike. They started making cheap comments and teasing me. It was late and the area was secluded. I got scared as they started revolving their bike around me. I started driving in the direction of a crowded place that is when they left. |
| Showing Pornography | A man in a car parked outside was watching porn when I saw him. He turned his device towards me and did offensive hand gestures inviting me. |
| Sexual Assault | When I was 7 years old, the shopkeeper removed my clothes and started touching me everywhere. He also tried to do it to another girl and failed. |
| Domestic Violence | My husband always doubts on my character and doesn't allow me to go outside alone, uses very vulgar language and beats me. |

**Table 1:** Examples of harassment cases shared on an incident reporting platform. Proactive action by authorities and citizens can help prevent numerous such incidents. Providing stakeholders with a concise overview of incidents occurring in a specific area is crucial and this can be effectively achieved by utilizing summarization algorithms.

In recent decades, the widespread availability of the internet has provided seamless access to online platforms to millions of people. Governments worldwide are increasingly utilizing these platforms to gather information directly from citizens – referred to as **Citizen Reporting** (Kopackova and Libalova 2019). By leveraging tools such as mobile applications, web-based portals, and social media integrations, citizen reporting platforms establish a direct and efficient communication link between individuals and the relevant authorities, enabling faster issue resolution and facilitating active public participation in community improvement. Beyond immediate problem solving, real-time data gathered through these platforms contributes valuable information for urban planning and proactive measures, paving the way for more efficient and adaptive communities. Citizen reporting typically addresses topics such as community issues, environmental challenges, crime prevention, public health, and disaster response (Shin et al. 2024). A special category of citizen reporting platforms, such as [Safe City](https://webapp.safecity.in), [SHe-Box](https://shebox.wcd.gov.in), and [JDoe](https://jdoe.io), allow people to post incidents of sexual harassment, domestic abuse, violence and assault. Table 1 showcases some example incidents shared by users on one such platform. While such horrific incidents cannot be entirely avoided through reporting alone, these incident reporting platforms can play a crucial role in preventing certain cases of sexual assault. By enabling users to analyze reported incidents, assess the safety of specific locations, and make informed decisions when traveling to potential hotspots, these platforms contribute to enhanced personal safety and awareness. Similarly, the local authorities can also benefit from these platforms to assess emerging cases, identify the underlying factors and determine proactive measures for effective resolution.

However, the challenge for the authorities is to navigate the high volume of information in such platforms. Manually reviewing all posts is often impractical, necessitating a summarization algorithm that can identify and select posts that are diverse as well as representative of the original data. Additionally, incident reporting platforms often feature a curated selection of posts on their homepage to showcase their core purpose, mission, and services. This deliberate selection also acts as a form of summarization. Summarization algorithms are of two types: 'extractive' and 'abstractive'. In **extractive summarization**, the algorithm selects a subset representative of the original text (Xu et al. 2020; Zhong et al. 2020; Zhang et al. 2022; Dash et al. 2019; Zhang et al. 2023a). In contrast, **abstractive summarization** algorithms generate summaries that capture the essence of the original text, often paraphrasing the content (Pu et al. 2023). For incident reporting platforms, extractive summarization is more suitable, as the goal is not to paraphrase the posts but to select a few that accurately capture a snapshot of the original content. When summarizing such sensitive posts, preserving the user's exact words is essential, making extractive summarization particularly valuable in maintaining authenticity and context. Several extractive summarization algorithms for user generated content have been proposed in the literature, primarily for text written in English (Bhattacharya et al. 2021; Kanwal and Rizzo 2022; Mukherjee et al. 2020; Jia et al. 2020). But there are several countries where English is not the primary language and users frequently communicate in code-mixed forms. For instance, India recognizes 22 official languages and users often post in Hinglish (a mix of Hindi and English). Such multilinguality limits the applicability of existing algorithms for extractive summarization of incident posts. In recent years, Large Language Models (LLMs) have demonstrated very good performance across various tasks in multilingual and code-mixed settings (Ouyang et al. 2022; Brown et al. 2020; Tang et al. 2023a; Jin et al. 2024a). Plus, summaries generated by LLMs showcase high coherence and are overwhelmingly preferred by human evaluators over other baseline algorithms (Pu et al. 2023; Liu et al. 2024). These prior results motivated us to investigate the utility of LLMs for extractive summarization of large volumes of user generated posts. However, we encountered two significant limitations which hinder the immediate application of LLMs for extractive summarization:

1. As generative models, LLMs perform abstractive summarization by paraphrasing rather than selecting the most relevant sentences (as shown in Figure 1) (Worledge et al. 2024).
2. Due to the finite size of the context window, LLMs cannot handle long texts in a single input, underscoring the need for a method that allows for processing long text (Jin et al. 2024b).

To overcome these limitations, in this paper, we present a novel framework **LaMSUM** (Large Language Model based Extractive SUMmarization) that integrates LLM-generated summaries with voting algorithms borrowed from Social Choice Theory (Brandt et al. 2016). Our judicious application of voting algorithms with a multi-level summarization framework ensures that LaMSUM outperforms the state-of-the-art fine-tuned summarization models. In summary, in this work, we make the following contributions:

- We propose a novel framework **LaMSUM** which can produce extractive summaries from large (having >>30K tokens) collection of user generated content. LaMSUM considers a multi-level summarization model that utilizes voting algorithms to combine LLM outputs to generate robust summaries.
- Extensive experiments with incident posts demonstrate that LaMSUM outperforms the state-of-the-art extractive summarization algorithms. To our knowledge, this is the first work to implement extractive summarization of a large collection of user-generated texts using LLMs by combining summarization with voting algorithms. At the same time, we demonstrate the effectiveness of such algorithms to facilitate data-driven decision-making promoting safer communities by providing actionable insights into reported incidents.

Code is available at https://github.com/garimachhikara128/LaMSUM

## 2 Background and Related Work

In this section, we review the relevant prior works that provide the foundation for our current research.

**Figure 1:** Current LLMs, by default, produce abstractive summaries. Llama-3.3-70b, despite specifically prompted for extractive summarization, generates abstractive summaries. This behavior underscores the need for a targeted approach to enable LLMs to effectively generate extractive summaries.

### AI Solution through Citizen Reporting

Web and social media platforms receive posts on sensitive issues such as online harassment, hate speech, abusive behavior, violence etc. Abuse experienced by users leads to mental stress often forcing them to leave the platform (Sambasivan et al. 2019; Thomas et al. 2022; Kim et al. 2024). Several AI-powered solutions have been designed to address solutions to these critical issues. Machine learning based classifiers and language models are utilized to detect the cases of sexual abuse, hate speech, offensive language, human trafficking and harassment cases (Sawhney et al. 2021; Hassan et al. 2020; Davidson et al. 2017; Singh et al. 2025; Upadhayaye et al. 2021; Stoop et al. 2019; Ghosh Chowdhury et al. 2019). Modelling the cyberbullying behavior using social network and language-based features can improve the classifier performance (Ziems et al. 2020; Olteanu et al. 2018). Development of a mobile computing based reporting tool empowers individuals with intellectual and developmental disabilities (I/DD) to self report abuse and share the incident with the intended group (Venkatasubramanian et al. 2021; Sultana et al. 2021). With the rise of public figures encouraging women to speak up, the number of non-anonymous self-reported assault stories has increased (ElSherief et al. 2017). Counterspeech is proven to be a viable alternative to blocking or suspending problematic messages or accounts, as it better aligns with the principles of free speech (Mathew et al. 2019). Conversational agents (CAs) have attracted significant interest as potential counselors due to their features, such as anonymity, which can help address many challenges associated with human-human interaction (Park and Lee 2021).

### Large Language Models (LLMs) for Summarization

LLMs are now being extensively used for summarization (Brown et al. 2020; Tang et al. 2023a; Jin et al. 2024a). Multiple works have proposed few-shot learning frameworks for the abstractive summarization of news, documents, webpages, and generic texts (Zhang et al. 2023b; Tang et al. 2023b; Yang et al. 2023; Bražinskas et al. 2020; Laskar et al. 2023), but their primary focus remains on short documents that can fit in the LLM context window. Researchers have also observed that human evaluators are increasingly preferring LLM-generated summaries compared to other baselines (Zhang et al. 2024; Wu et al. 2024; Goyal et al. 2023; Zhang et al. 2023c; Liu et al. 2024). Despite the advancements, recent studies have also uncovered factual inaccuracies and inconsistencies in LLM-generated summaries (Tang et al. 2024; Tam et al. 2023; Luo et al. 2023; Laban et al. 2023).

**Figure 2:** LaMSUM: Multi-level framework for extractive summarization of large user-generated text. Input set 𝒯 (level 0) is divided into ⌈|