A Deep Investigation on News Aggregation and Recommendation System: NARS

Recent rises in political polarization across the globe are often ascribed to algorithmic content filtering on social media, news platforms, or search engines. The widespread usage of news recommendation systems (NRS) is theorized to drive users in homogenous information environments and, thereby, drive affective, ideological, and perceived polarization. To test this assumption, we conducted an online experiment ( n = 750) with running algorithms that enriches content-based NRS with negative or neutral sentiment. Our experiment finds only limited evidence for polarization effects of content-based NRS. Nevertheless, the time spent with an NRS and its recommended articles seems to play a crucial role as a moderator of polarization. The longer participants were using an NRS enriched with negative sentiment, the more they got affectively polarized, whereas participants using an NRS incorporating balanced sentiment ideologically depolarized over time. Implications for future research are discussed.

This study aims to explore the impact of personalized news recommendation systems based on big data and artificial intelligence algorithms on audience cognition and behavior. With the rapid development of information technology, news recommendation systems have been widely used in the media industry, but their impact on audiences has not been fully studied. This paper adopts a mixed research method, combining questionnaire surveys, in-depth interviews and experimental research, and conducts a three-month follow-up survey on 500 news app users. The results show that the personalized recommendation system significantly increases users' news reading volume and retention time, but also leads to the intensification of information cocoon effect and cognitive bias. In addition, the study finds that the transparency of the recommendation algorithm, the user's media literacy level and the diversity of news sources are key factors affecting audience cognition and behavior. Based on the research findings, this paper proposes suggestions for optimizing the news recommendation system, improving users' media literacy and promoting information diversification, in order to minimize the negative impact of personalized recommendations while improving user experience. This study has important theoretical and practical significance for understanding the characteristics of audience behavior in the intelligent media environment, improving the design of news recommendation systems and formulating relevant policies.

How do users interact with algorithm recommender systems? The interaction of users, algorithms, and performance

Recommender systems are becoming an essential part of smart services. When building a news recommender system, we should consider special features different from other recommender systems. Hot news topics are changing every moment, thus it is important to recommend right news at the right time. This paper aims to propose a new model, based on deep neural network, to analyse user preference for news recommender system. The model extracts interest keywords to characterize the user preference from the set of news articles read by that particular user in the past. The model utilizes characterizing features for news recommendation, and applies those to the keyword classification for user preference. For the keyword classification, we use deep neural network for online preference analysis, because adaptive learning is necessary to track changes of hot topics sensitively. The usefulness of our model is validated through experiments. In addition, the accuracy and diversity of the recommendation results is also analysed.

Using complex language models has been a common strategy as personalized news recommendation systems are adopted by online news sites more and more. Before the epoch of GPT-3, news recommendation systems underwent a progression from rule-based and collaborative filtering approaches in the pre-2010s, through the integration of neural networks in the 2010s, to the emergence of earlier iterations of large language models like GPT-2 in 2019. Pre-trained language models have ushered in a new era of recommendation paradigms, thanks to the emergence of huge language models like GPT-3 and T-5. With its easy-to-use interface, ChatGPT is becoming more and more popular for text-based jobs. Focusing on news provider fairness, individualized news recommendations, and fake news identification, this study starts an inquiry into ChatGPT's efficacy in news recommendations. We acknowledge that ChatGPT's output sensitivity to input phrasing is a limitation, and our goal is to investigate these limitations from each angle. We also investigate whether certain prompt formats can help to alleviate these constraints or if more research is needed. To go beyond set assessments, we create a webpage where ChatGPT's performance on the examined activities and prompts is tracked once a week. Through the use of big language models, this work seeks to improve news recommendation performance and stimulate more research in this area.

Recommender systems are widely applied in digital platforms such as news websites to personalize services based on user preferences. In news websites most of users are anonymous and the only available data is sequences of items in anonymous sessions. Due to this, typical collaborative filtering methods, which are highly applied in many applications, are not effective in news recommendations. In this context, session-based recommenders are able to recommend next items given the sequence of previous items in the active session. Neighborhood-based session-based recommenders has been shown to be highly effective compared to more sophisticated approaches. In this study we propose scenarios to make these session-based recommender systems diversity-aware and to address the filter bubble phenomenon. The filter bubble phenomenon is a common concern in news recommendation systems and it occurs when the system narrows the information and deprives users of diverse information. The results of applying the proposed scenarios show that these diversification scenarios improve the diversity measures in these session-based recommender systems based on four news datasets.

News recommender systems rely on rich textual information from news articles to generate user-specific recommendations. This reliance may expose these systems to potential vulnerabilities through textual attacks. To explore this vulnerability, we propose LANCE, a LArge language model-based News Content rEwriting framework, designed to influence news rankings and highlight the unintended promotion of manipulated news. LANCE consists of two key components: an explorer and a reflector. The explorer first generates rewritten news using diverse prompts, incorporating different writing styles, sentiments, and personas. We then collect these rewrites, evaluate their ranking impact within news recommender systems, and apply a filtering mechanism to retain effective rewrites. Next, the reflector fine-tunes an open-source LLM using the successful rewrites, enhancing its ability to generate more effective textual attacks. Experimental results demonstrate the effectiveness of LANCE in manipulating rankings within news recommender systems. Unlike attacks in other recomendation domains, negative and neutral rewrites consistently outperform positive ones, revealing a unique vulnerability specific to news recommendation. Once trained, LANCE successfully attacks unseen news recommender systems (i.e., those for which LANCE received no information during training), highlighting its generalization ability and exposing shared vulnerabilities across different systems. Our work underscores the urgent need for research on textual attacks and paves the way for future studies on defense strategies.

Recommender systems are designed to help users quickly access large volumes of information according to their profiles. Most previous works in recommender systems have put their emphasis on the accuracy of finding the most similar items according to a user’s profile, while often ignoring other aspects that may affect users’ experiences in practice, e.g., the novelty and diversity issues within a recommendation list. In this paper, we focus on utilizing taxonomic knowledge extracted from an online encyclopedia to boost a content-based personalized news recommender system without much human involvement. Given a recommendation list, we improve a user’s satisfaction by introducing the taxonomy based novelty and diversity metrics to include novel, but potentially related items into the list, and filter out redundant ones. The experimental results show that the coarse grained knowledge resources can help a content-based news recommender system provides accurate as well as user-oriented recommendations.

Numerous recommendation approaches are in use today. However, comparing their effectiveness is a challenging task because evaluation results are rarely reproducible. In this article, we examine the challenge of reproducibility in recommender-system research. We conduct experiments using Plista's news recommender system, and Docear's research-paper recommender system. The experiments show that there are large discrepancies in the effectiveness of identical recommendation approaches in only slightly different scenarios, as well as large discrepancies for slightly different approaches in identical scenarios. For example, in one news-recommendation scenario, the performance of a content-based filtering approach was twice as high as the second-best approach, while in another scenario the same content-based filtering approach was the worst performing approach. We found several determinants that may contribute to the large discrepancies observed in recommendation effectiveness. Determinants we examined include user characteristics (gender and age), datasets, weighting schemes, the time at which recommendations were shown, and user-model size. Some of the determinants have interdependencies. For instance, the optimal size of an algorithms' user model depended on users' age. Since minor variations in approaches and scenarios can lead to significant changes in a recommendation approach's performance, ensuring reproducibility of experimental results is difficult. We discuss these findings and conclude that to ensure reproducibility, the recommender-system community needs to (1) survey other research fields and learn from them, (2) find a common understanding of reproducibility, (3) identify and understand the determinants that affect reproducibility, (4) conduct more comprehensive experiments, (5) modernize publication practices, (6) foster the development and use of recommendation frameworks, and (7) establish best-practice guidelines for recommender-systems research.

Increasingly, algorithms play an important role in everyday decision-making processes. Recommender systems, specifically, are algorithms that serve to influence end-users’ decision-making (e.g. what to read, who to befriend, who to rent to…). However, the companies that develop and produce these systems are not neutral, but have an economic goal and specific vision on how society should operate. These algorithms should thus never be trusted blindly.
 An algorithm consists of collective human practices and consequently warm human and institutional choices. Therefore, they should be perceived as culture. Despite the many academics that are joining the debate to denounce the bias, opaqueness and unfairness often found in these algorithms, little empirical research has invested in treating algorithms in its socio-technical assembly as culture.
 To better understand how end-users perceive these algorithmic systems, we strive to understand how they imagine and (dis)trust the different components of the socio-technical assembly. We are demystifying the imagined processes incorporated in these algorithmic systems in the minds of the end-user using a deconstructed version of Buchers’ (2017) algorithmic imaginary.
 Currently, companies put ever more effort into personalizing news, using news recommender systems (NRS). NRS organize, select and aggregate news to influence the decision-making of an end-user without a transparent explanation on the process. Therefore, we focus our study on the end-users of these NRS.
 In this qualitative study, we are interviewing 25 end-users of NRS to understand the assumptions and apprehend the (dis-)trust people have about the different elements of the socio-technical assembly of news recommender systems.

In recent years, the type and quantity of news are growing rapidly, and it is not easy for users to find the news they are interested in the massive amount of news. A news recommendation system can score and predict the candidate news, and finally recommend the news with high scores to users. However, existing user models usually only consider users’ long-term interests and ignore users’ recent interests, which affects users’ usage experience. Therefore, this paper introduces gated recurrent unit (GRU) sequence network to capture users’ short-term interests and combines users’ short-term interests and long-term interests to characterize users. While existing models often only use the user’s browsing history and ignore the variability of different users’ interest in the same news, we introduce additional user’s ID information and apply the personalized attention mechanism for user representation. Thus, we achieve a more accurate user representation. We also consider the risk of compromising user privacy if the user model training is placed on the server side. To solve this problem, we design the training of the user model locally on the client side by introducing a federated learning framework to keep the user’s browsing history on the client side. We further employ secure multi-party computation to request news representations from the server side, which protects privacy to some extent. Extensive experiments on a real-world news dataset show that our proposed news recommendation model has a better improvement in several performance evaluation metrics. Compared with the current state-of-the-art federated news recommendation models, our model has increased by 0.54% in AUC, 1.97% in MRR, 2.59% in nDCG@5%, and 1.89% in nDCG@10. At the same time, because we use a federated learning framework, compared with other centralized news recommendation methods, we achieve privacy protection for users.

The most current news recommendations are suitable for news which comes from a single news website, not for news from different heterogeneous news websites. Previous researches about news recommender systems based on different strategies have been proposed to provide news personalization services for online news readers. However, little research work has been reported on utilizing hundreds of heterogeneous news websites to provide top hot news services for group customers (e.g., government staffs). In this paper, we propose a hot news recommendation model based on Bayesian model, which is from hundreds of different news websites. In the model, we determine whether the news is hot news by calculating the joint probability of the news. We evaluate and compare our proposed recommendation model with the results of human experts on the real data sets. Experimental results demonstrate the reliability and effectiveness of our method. We also implement this model in hot news recommendation system of Hangzhou city government in year 2013, which achieves very good results.

With the continuous development of information technology, we are now in an era of information explosion, and we are also facing the problem of information overload, i.e., it is difficult to find the news that interests us from the huge news sources on the Internet. Based on this problem, this paper proposes a perceptron model-based news recommendation system, which uses the perceptron model to extract news keywords and realize the classification of the news; for user interest preferences, we analyze users’ news browsing preferences by collecting user behavior logs and use a multi-layer perceptron model to aggregate users’ interest features. The perceptron model is compared with traditional recommendation methods in terms of accuracy and recall, and one user is selected from 50 users who have been using the system for three months to analyze their news browsing preferences by single-day, short-term and long-term news data to verify the effectiveness of the perceptron model in aggregating users’ interest features. In terms of accuracy and recall, the accuracy of the perceptron model is higher than that of the traditional method for each news type, improving the accuracy by at least 5%-10%, e.g., 10% for real estate news and 9% for financial news. The recall rate of the perceptron model is improved by more than 6% for all news types except for social and entertainment news types. In terms of user interest preference mining, the whole change process of user browsing preference trend is in line with the human forgetting curve, which verifies that the perceptron model in the system is better used and meets the design requirements. Therefore, the applicability and reliability of the model proposed in this paper for news recommendation systems are verified in terms of accuracy, recall, and user interest preference mining.

Ubiquitous access to internet has resulted in more and more people going online to get their daily dose of news. In a 2010 survey conducted by the Pew Project for Excellence in Journalism, 41% of the respondents said they get most of their news online, 10% more than those who said they got most of their news from a newspaper. A lot of socio-technical factors have contributed to this phenomenal rise in adoption of online news in recent years. One of the biggest reasons why people are increasingly reading news online is because it facilitates discussion with peers (Nguyen 2010), offering different viewpoints which aid in forming a rounded personal opinion about the news story. The Pew survey found that 37% of online news users (and 51% of 18-29 year olds) think that commenting on news stories is an important feature to have. A lot of people tend to shape their opinion by reading discussion comments, reflective articles, blogs and even tweets about the news. Hence, an increasing number of people rely on online sources of news – be it news websites or news aggregator services like Digg, Reddit, Google Reader, Flipboard, Pulse etc. The problem with these news websites and aggregators is that the only way people can gather public opinion is by actively searching through the endless stream of comments and feeds, filtering out spam (which is a growing problem) and then reading the relevant posts. A top trending story on Twitter will typically see multiple tweets per second, and keeping up with the rapid flow of incoming tweets is quite cumbersome and cognitively taxing. Hence it becomes increasingly difficult and time consuming for someone who wants to get the pulse of the people affected by a news story. Furthermore, in certain scenarios people might want to look at more fine grained opinions. Currently, there is no elegant way to extract geographic and demographic impact of a news story. What is the public sentiment in Indonesia about the Arab Spring? How did the public opinion about the Wikileaks disclosures change as the story unfolded during the course of a year? It is very difficult and tedious to observe such patterns using the currently available news providers. This work attempts to solve these problems by proposing a news aggregator platform which pulls news stories from various sources and also aggregates public responses, reflections, opinions and sentiments associated with those stories. This data is presented in ways that are easily understandable so readers can make better sense of the stories unfolding across the globe. Such a news aggregator platform that gathers and display public opinion and sentiments about a story, must deal with various challenges – 1. Opinions are very subjective. Different people feel about a story in different ways. With such an enormous amount of diverse opinions and subjectivity, how can we possibly aggregate the responses into something that makes sense as a whole? ________________________________ Acknowledgements: Prof. Yardi, S. for the guidance Sethi, P. (2013). Public opinion aggregation by annotation and tagging of online news stories. iConference 2013 Proceedings (pp. 891-894). doi:10.9776/13453 Copyright is held by the author. iConference 2013 February 12-15, 2013 Fort Worth, TX, USA 892 2. There isn’t really a unified web standard for expressing opinion (in textual form). Some people tweet in 140 characters, while others write elaborate blog posts. Some websites employ tags which a reader can use to define and classify their public opinion, while others rely on threaded comments and comment ranking systems. How can a platform be flexible enough to adapt to all these varied standards so that it can extract valuable data from various sources? Perhaps the platform can create a new standard of expression on the web which is flexible and comprehensive enough to be used to express diverse views about every news story in the world. 3. How to filter out spam while extracting public opinion? 4. Once the platform has access to the data it needs, how should it be displayed to the reader in a way that makes sense? What forms of visualizations, illustrations and graphical representations can be employed to give the reader a holistic view of how people feel about a story? 5. How can the platform determine and convey effects of geographical, demographic and temporal variations as the story unfolds? These are just a few out of possibly many issues which must be dealt with. Previous research on similar public opinion aggregation services has greatly focused on natural language processing, data mining and text categorization and clustering. Xiaojun (2010) proposed a framework for crawling the web for comments and applying various data mining algorithms on the data to extract relevant information. Diakopoulos and Shamma (2010) used tweets posted in conjunction with the live presidential debate between Barack Obama and John McCain to gauge public opinion. Brody and Diakopoulos (2011) studied the use of word lengthening to detect sentiment in microblogs. This research proposes a solution – The Opinionated Reader, which relies on sentiment tags and annotations associated with a news story. The essential idea is to create a commenting, discussion and sharing plug-in which can be used by news websites and aggregators as a commenting solution for their news pages. Users wanting to share or comment on a news story through the plug-in are asked to tag the news story with sentiment tags and annotate the story with their reaction (happy/positive or sad/negative). These tags and annotations are stored, aggregated and linked to each news story. A mobile application provides the front-end interface for users to access the news stories and the aggregated sentiment associated with each story. The basic architecture is explained as follows: The Opinionated Reader – Mobile/Tablet App The app fetches news articles from various web sources, based on the interests and preferences configured by the user. In every news article, a portion of the screen real estate is reserved for Opinions which shows graphical visualizations and illustrations of the public opinion surrounding the news story. These visualizations include: A Sentiment Graph indicating the popular tags associated with the story (E.g.: “Shocking”, “Inspiring”, “Amusing” etc.). See Figure 1 for example visualization. A Positivity Graph which plots the level of positivity associated with the story on a time scale from when the news broke. See Figure 2 for example visualization. Figure 1. Example Sentiment Graph Figure 2. Example Positivity Graph iConference 2013 February 12-15, 2013 Fort Worth, TX, USA 893 The user can choose to see these visualizations for a particular time period in the evolution history of the news story, or for a specific country. The app also facilitates people to tag and annotate news articles from within its interface. The Opinionated Reader – Commenting and Sharing Web Plug-in These days, a common way of adding discussion and commenting functionality to news websites is by using 3 rd party services (like DISQUS). The Opinionated Reader is a similar service which can be embedded into the news articles of various news websites to enable commenting and sharing. When someone wishes to comment on an article, the comment is directed through this plug-in, which allows the users to annotate the article with the sentiment tags and reactions along with their comments. The Opinionated Reader saves this information along with the commenter’s location and date of comment (See Figure 3). Figure 3. The Opinionated Reader Web Plug-in 'Add Comment' dialog mockup The Opinionated Reader – Back-end The Back-end maintains a database of news items extracted from RSS feeds of various news websites. Each news article is linked with the sentiment tags and reaction/positivity annotations extracted from the comments and annotations gathered by the commenting plug-in. This data is used by the mobile/tablet app to generate visualizations (Sentiment Graph and Positivity Graph). The back-end also performs data mining on the tags and annotations for geographies and tracks the opinions across time. The back end system responds to queries received from the mobile app with the news story and associated tags and annotations, which are then rendered by the mobile app for the user. iConference 2013 February 12-15, 2013 Fort Worth, TX, USA 894 Discussion and Conclusion This design idea is still in a nascent state and has long hours of research, brainstorming, designing and development to go before it can be realized into something tangible. Twitter has grown exponentially in importance as a news source and it would be vastly valuable to integrate Twitter with The Opinionated Reader. Possibilities include use of special hash tags and natural language processing of tweets to extract public sentiment. The current design supports only two reaction annotations – positive and negative. Not every news story fits this annotation paradigm. Further research about human reactions to news stories might unveil interesting insights which would help zero in on a more robust annotation rubric. Lastly, since this service is envisioned to be non-curated and non-moderated, the value served by the app depends on the users themselves. Greater adoption will lead to more annotations and tags, which translates into a more accurate public opinion as presented to the user.

A concept is a key phrase which can represent an entity, event or idea that people are interested in. Concept-centric Web news recommendation is a novel content-based recommendation paradigm which can partially alleviate the cold-start problem and provide better recommendation results in terms of diversity than traditional news recommendation systems, as it can capture users’ interest in a natural way and can even recommend a new Web news to a user as long as it is conceptually relevant to a main concept of the Web news the user is browsing. This demonstration paper presents a novel CONcept-Centric nEws Recommendation sysTem called CONCERT. CONCERT consists of two parts: (1) A concept extractor which is based on machine learning algorithms and can extract main concepts from Web news pages, (2) A real-time recommender which recommends conceptually relevant Web news to a user based on the extracted concepts.