Yiping Jin

Yiping Jin, Vishakha Kadam, Dittaya Wanvarie

Pre-trained language models (PLMs) fail to generate long-form narrative text because they do not consider global structure. As a result, the generated texts are often incohesive, repetitive, or lack content. Recent work in story generation reintroduced explicit content planning in the form of prompts, keywords, or semantic frames. Trained on large parallel corpora, these models can generate more logical event sequences and thus more contentful stories. However, these intermediate representations are often not in natural language and cannot be utilized by PLMs without fine-tuning. We propose generating story plots using off-the-shelf PLMs while maintaining the benefit of content planning to generate cohesive and contentful stories. Our proposed method, ScratchPlot, first prompts a PLM to compose a content plan. Then, we generate the story's body and ending conditioned on the content plan. Furthermore, we take a generate-and-rank approach by using additional PLMs to rank the generated (story, ending) pairs. We benchmark our method with various baselines and achieved superior results in both human and automatic evaluation.

Yiping Jin, Leo Wanner, Alexander Shvets

Online hate detection suffers from biases incurred in data sampling, annotation, and model pre-training. Therefore, measuring the averaged performance over all examples in held-out test data is inadequate. Instead, we must identify specific model weaknesses and be informed when it is more likely to fail. A recent proposal in this direction is HateCheck, a suite for testing fine-grained model functionalities on synthesized data generated using templates of the kind "You are just a [slur] to me." However, despite enabling more detailed diagnostic insights, the HateCheck test cases are often generic and have simplistic sentence structures that do not match the real-world data. To address this limitation, we propose GPT-HateCheck, a framework to generate more diverse and realistic functional tests from scratch by instructing large language models (LLMs). We employ an additional natural language inference (NLI) model to verify the generations. Crowd-sourced annotation demonstrates that the generated test cases are of high quality. Using the new functional tests, we can uncover model weaknesses that would be overlooked using the original HateCheck dataset.

Chen Huang, Yiping Jin, Ilija Ilievski et al.

Human annotation is a time-consuming task that requires a significant amount of effort. To address this issue, interactive data annotation utilizes an annotation model to provide suggestions for humans to approve or correct. However, annotation models trained with limited labeled data are prone to generating incorrect suggestions, leading to extra human correction effort. To tackle this challenge, we propose Araida, an analogical reasoning-based approach that enhances automatic annotation accuracy in the interactive data annotation setting and reduces the need for human corrections. Araida involves an error-aware integration strategy that dynamically coordinates an annotation model and a k-nearest neighbors (KNN) model, giving more importance to KNN's predictions when predictions from the annotation model are deemed inaccurate. Empirical studies demonstrate that Araida is adaptable to different annotation tasks and models. On average, it reduces human correction labor by 11.02% compared to vanilla interactive data annotation methods.

Yiping Jin, Leo Wanner, Aneesh Moideen Koya

Hate speech (HS) classifiers do not perform equally well in detecting hateful expressions towards different target identities. They also demonstrate systematic biases in predicted hatefulness scores. Tapping on two recently proposed functionality test datasets for HS detection, we quantitatively analyze the impact of different factors on HS prediction. Experiments on popular industrial and academic models demonstrate that HS detectors assign a higher hatefulness score merely based on the mention of specific target identities. Besides, models often confuse hatefulness and the polarity of emotions. This result is worrisome as the effort to build HS detectors might harm the vulnerable identity groups we wish to protect: posts expressing anger or disapproval of hate expressions might be flagged as hateful themselves. We also carry out a study inspired by social psychology theory, which reveals that the accuracy of hatefulness prediction correlates strongly with the intensity of the stereotype.

Yiping Jin, Leo Wanner, Vishakha Laxman Kadam et al.

As pointed out by several scholars, current research on hate speech (HS) recognition is characterized by unsystematic data creation strategies and diverging annotation schemata. Subsequently, supervised-learning models tend to generalize poorly to datasets they were not trained on, and the performance of the models trained on datasets labeled using different HS taxonomies cannot be compared. To ease this problem, we propose applying extremely weak supervision that only relies on the class name rather than on class samples from the annotated data. We demonstrate the effectiveness of a state-of-the-art weakly-supervised text classification model in various in-dataset and cross-dataset settings. Furthermore, we conduct an in-depth quantitative and qualitative analysis of the source of poor generalizability of HS classification models.

Yiping Jin, Vishakha Kadam, Dittaya Wanvarie

Contextual advertising provides advertisers with the opportunity to target the context which is most relevant to their ads. However, its power cannot be fully utilized unless we can target the page content using fine-grained categories, e.g., "coupe" vs. "hatchback" instead of "automotive" vs. "sport". The widely used advertising content taxonomy (IAB taxonomy) consists of 23 coarse-grained categories and 355 fine-grained categories. With the large number of categories, it becomes very challenging either to collect training documents to build a supervised classification model, or to compose expert-written rules in a rule-based classification system. Besides, in fine-grained classification, different categories often overlap or co-occur, making it harder to classify accurately. In this work, we propose wiki2cat, a method to tackle the problem of large-scaled fine-grained text classification by tapping on Wikipedia category graph. The categories in IAB taxonomy are first mapped to category nodes in the graph. Then the label is propagated across the graph to obtain a list of labeled Wikipedia documents to induce text classifiers. The method is ideal for large-scale classification problems since it does not require any manually-labeled document or hand-curated rules or keywords. The proposed method is benchmarked with various learning-based and keyword-based baselines and yields competitive performance on both publicly available datasets and a new dataset containing more than 300 fine-grained categories.