A conspiracy theory (CT) suggests covert groups or powerful individuals secretly manipulate events. Not knowing about existing conspiracy theories could make one more likely to believe them, so this work aims to compile a list of CTs shaped as a tree that is as comprehensive as possible. We began with a manually curated ‘tree’ of CTs from academic papers and Wikipedia. Next, we examined 1769 CT-related articles from four fact-checking websites, focusing on their core content, and used a technique called Keyphrase Extraction to label the documents. This process yielded 769 identified conspiracies, each assigned a label and a family name. The second goal of this project was to detect whether an article is a conspiracy theory, so we built a binary classifier with our labeled dataset. This model uses a transformer-based machine learning technique and is pre-trained on a large corpus called RoBERTa, resulting in an F1 score of 87%. This model helps to identify potential conspiracy theories in new articles. We used a combination of clustering (HDBSCAN) and a dimension reduction technique (UMAP) to assign a label from the tree to these new articles detected as conspiracy theories. We then labeled these groups accordingly to help us match them to the tree. These can lead us to detect new conspiracy theories and expand the tree using computational methods. We successfully generated a tree of conspiracy theories and built a pipeline to detect and categorize conspiracy theories within any text corpora. This pipeline gives us valuable insights through any databases formatted as text.
- Award ID(s):
- 1750978
- NSF-PAR ID:
- 10091995
- Date Published:
- Journal Name:
- Annual Conference of the North American Chapter of the Association for Computational Linguistics
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract -
more » « less
Abstract This article examines the automatic identification of Web registers, that is, text varieties such as news articles and reviews. Most studies have focused on corpora restricted to include only preselected classes with well-defined characteristics. These corpora feature only a subset of documents found on the unrestricted open Web, for which register identification has been particularly difficult because the range of linguistic variation on the Web is known to be substantial. As part of this study, we present the first open release of the Corpus of Online Registers of English (CORE), which is drawn from the unrestricted open Web and, currently, is the largest collection of manually annotated Web registers. Furthermore, we demonstrate that the CORE registers can be automatically identified with competitive results, with the best performance being an F1-score of 68% with the deep learning model BERT. The best performance was achieved using two modeling strategies. The first one involved modeling the registers using propagated register labels, that is, repeating the main register label along with its corresponding subregister label in a multilabel model. In the second one, we explored how the length of the document affects model performance, discovering that the beginning provided superior classification accuracy. Overall, the current study presents a systematic approach for the automatic identification of a large number of Web registers from the unrestricted Web, hence providing new pathways for future studies.
-
more » « less
Abstract Journals play a critical role in the scientific process because they evaluate the quality of incoming papers and offer an organizing filter for search. However, the role of journals has been called into question because new preprint archives and academic search engines make it easier to find articles independent of the journals that publish them. Research on this issue is complicated by the deeply confounded relationship between article quality and journal reputation. We present an innovative proxy for individual article quality that is divorced from the journal's reputation or impact factor: the number of citations to preprints posted on
arXiv.org . Using this measure to study three subfields of physics that were early adopters of arXiv, we show that prior estimates of the effect of journal reputation on an individual article's impact (measured by citations) are likely inflated. While we find that higher‐quality preprints in these subfields are now less likely to be published in journals compared to prior years, we find little systematic evidence that the role of journal reputation on article performance has declined. -
With the spread of the SARS-CoV-2, enormous amounts of information about the pandemic are disseminated through social media platforms such as Twitter. Social media posts often leverage the trust readers have in prestigious news agencies and cite news articles as a way of gaining credibility. Nevertheless, it is not always the case that the cited article supports the claim made in the social media post. We present a cross-genre ad hoc pipeline to identify whether the information in a Twitter post (i.e., a “Tweet”) is indeed supported by the cited news article. Our approach is empirically based on a corpus of over 46.86 million Tweets and is divided into two tasks: (i) development of models to detect Tweets containing claim and worth to be fact-checked and (ii) verifying whether the claims made in a Tweet are supported by the newswire article it cites. Unlike previous studies that detect unsubstantiated information by post hoc analysis of the patterns of propagation, we seek to identify reliable support (or the lack of it) before the misinformation begins to spread. We discover that nearly half of the Tweets (43.4%) are not factual and hence not worth checking – a significant filter, given the sheer volume of social media posts on a platform such as Twitter. Moreover, we find that among the Tweets that contain a seemingly factual claim while citing a news article as supporting evidence, at least 1% are not actually supported by the cited news, and are hence misleading.more » « less
-
Obeid, I. ; Selesnik, I. ; Picone, J. (Ed.)The Neuronix high-performance computing cluster allows us to conduct extensive machine learning experiments on big data [1]. This heterogeneous cluster uses innovative scheduling technology, Slurm [2], that manages a network of CPUs and graphics processing units (GPUs). The GPU farm consists of a variety of processors ranging from low-end consumer grade devices such as the Nvidia GTX 970 to higher-end devices such as the GeForce RTX 2080. These GPUs are essential to our research since they allow extremely compute-intensive deep learning tasks to be executed on massive data resources such as the TUH EEG Corpus [2]. We use TensorFlow [3] as the core machine learning library for our deep learning systems, and routinely employ multiple GPUs to accelerate the training process. Reproducible results are essential to machine learning research. Reproducibility in this context means the ability to replicate an existing experiment – performance metrics such as error rates should be identical and floating-point calculations should match closely. Three examples of ways we typically expect an experiment to be replicable are: (1) The same job run on the same processor should produce the same results each time it is run. (2) A job run on a CPU and GPU should produce identical results. (3) A job should produce comparable results if the data is presented in a different order. System optimization requires an ability to directly compare error rates for algorithms evaluated under comparable operating conditions. However, it is a difficult task to exactly reproduce the results for large, complex deep learning systems that often require more than a trillion calculations per experiment [5]. This is a fairly well-known issue and one we will explore in this poster. Researchers must be able to replicate results on a specific data set to establish the integrity of an implementation. They can then use that implementation as a baseline for comparison purposes. A lack of reproducibility makes it very difficult to debug algorithms and validate changes to the system. Equally important, since many results in deep learning research are dependent on the order in which the system is exposed to the data, the specific processors used, and even the order in which those processors are accessed, it becomes a challenging problem to compare two algorithms since each system must be individually optimized for a specific data set or processor. This is extremely time-consuming for algorithm research in which a single run often taxes a computing environment to its limits. Well-known techniques such as cross-validation [5,6] can be used to mitigate these effects, but this is also computationally expensive. These issues are further compounded by the fact that most deep learning algorithms are susceptible to the way computational noise propagates through the system. GPUs are particularly notorious for this because, in a clustered environment, it becomes more difficult to control which processors are used at various points in time. Another equally frustrating issue is that upgrades to the deep learning package, such as the transition from TensorFlow v1.9 to v1.13, can also result in large fluctuations in error rates when re-running the same experiment. Since TensorFlow is constantly updating functions to support GPU use, maintaining an historical archive of experimental results that can be used to calibrate algorithm research is quite a challenge. This makes it very difficult to optimize the system or select the best configurations. The overall impact of all of these issues described above is significant as error rates can fluctuate by as much as 25% due to these types of computational issues. Cross-validation is one technique used to mitigate this, but that is expensive since you need to do multiple runs over the data, which further taxes a computing infrastructure already running at max capacity. GPUs are preferred when training a large network since these systems train at least two orders of magnitude faster than CPUs [7]. Large-scale experiments are simply not feasible without using GPUs. However, there is a tradeoff to gain this performance. Since all our GPUs use the NVIDIA CUDA® Deep Neural Network library (cuDNN) [8], a GPU-accelerated library of primitives for deep neural networks, it adds an element of randomness into the experiment. When a GPU is used to train a network in TensorFlow, it automatically searches for a cuDNN implementation. NVIDIA’s cuDNN implementation provides algorithms that increase the performance and help the model train quicker, but they are non-deterministic algorithms [9,10]. Since our networks have many complex layers, there is no easy way to avoid this randomness. Instead of comparing each epoch, we compare the average performance of the experiment because it gives us a hint of how our model is performing per experiment, and if the changes we make are efficient. In this poster, we will discuss a variety of issues related to reproducibility and introduce ways we mitigate these effects. For example, TensorFlow uses a random number generator (RNG) which is not seeded by default. TensorFlow determines the initialization point and how certain functions execute using the RNG. The solution for this is seeding all the necessary components before training the model. This forces TensorFlow to use the same initialization point and sets how certain layers work (e.g., dropout layers). However, seeding all the RNGs will not guarantee a controlled experiment. Other variables can affect the outcome of the experiment such as training using GPUs, allowing multi-threading on CPUs, using certain layers, etc. To mitigate our problems with reproducibility, we first make sure that the data is processed in the same order during training. Therefore, we save the data from the last experiment and to make sure the newer experiment follows the same order. If we allow the data to be shuffled, it can affect the performance due to how the model was exposed to the data. We also specify the float data type to be 32-bit since Python defaults to 64-bit. We try to avoid using 64-bit precision because the numbers produced by a GPU can vary significantly depending on the GPU architecture [11-13]. Controlling precision somewhat reduces differences due to computational noise even though technically it increases the amount of computational noise. We are currently developing more advanced techniques for preserving the efficiency of our training process while also maintaining the ability to reproduce models. In our poster presentation we will demonstrate these issues using some novel visualization tools, present several examples of the extent to which these issues influence research results on electroencephalography (EEG) and digital pathology experiments and introduce new ways to manage such computational issues.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
