DNA Methylation Data to Predict Suicidal and Non-Suicidal Deaths: A Machine Learning Approach

Zahan, Rifat; McQuillan, Ian; Osgood, Nathaniel D.

doi:10.1109/ichi.2018.00057

“…[1] DNA Methylation Data to Predict Suicidal and Non-Suicidal Deaths: A ML. Approach [67] Semi-Supervised and Transfer Learning Approaches for Low Resource Senti. Class.…”

Section: Discussionmentioning

confidence: 99%

“…However, in contrast to the seed, GANs are not used for augmentation, and the classification task is not concerned with sentiment. The method-specific recommendations Zahan et al [67], Husmann et al [26], and Shen et al [56] are at first sight unrelated to the seed since they focus on unrelated topics such as hashing or the classification of biomedical or financial data. Nonetheless, the seed and the method-specific recommendation all use t-distributed Stochastic Neighbor Embedding (t-SNE) for visualization.…”

Section: Qualitative Verificationmentioning

confidence: 99%

Specialized Document Embeddings for Aspect-based Similarity of Research Papers

Ostendorff¹,

Blume²,

Ruas³

et al. 2022

Preprint

View full text Add to dashboard Cite

Document embeddings and similarity measures underpin contentbased recommender systems, whereby a document is commonly represented as a single generic embedding. However, similarity computed on single vector representations provides only one perspective on document similarity that ignores which aspects make two documents alike. To address this limitation, aspect-based similarity measures have been developed using document segmentation or pairwise multi-class document classification. While segmentation harms the document coherence, the pairwise classification approach scales poorly to large scale corpora. In this paper, we treat aspect-based similarity as a classical vector similarity problem in aspect-specific embedding spaces. We represent a document not as a single generic embedding but as multiple specialized embeddings. Our approach avoids document segmentation and scales linearly w.r.t. the corpus size. In an empirical study, we use the Papers with Code corpus containing 157, 606 research papers and consider the task, method, and dataset of the respective research papers as their aspects. We compare and analyze three generic document embeddings, six specialized document embeddings and a pairwise classification baseline in the context of research paper recommendations. As generic document embeddings, we consider FastText, SciBERT, and SPECTER. To compute the specialized document embeddings, we compare three alternative methods inspired by retrofitting, fine-tuning, and Siamese networks. In our experiments, Siamese SciBERT achieved the highest scores. Additional analyses indicate an implicit bias of the generic document embeddings towards the dataset aspect and against the method aspect of each research paper. Our approach of aspect-based document embeddings mitigates potential risks arising from implicit biases by making them explicit. This can, for example, be used for more diverse and explainable recommendations.

show abstract

“…However, in contrast to the seed, GANs are not used for augmentation, and the classification task is not concerned with sentiment. The method-specific recommendations Zahan et al [76], Husmann et al [29], and Shen et al [65] are at first sight unrelated to the seed since they focus on unrelated topics such as hashing or the classification of biomedical or financial data. Nonetheless, the seed and the method-specific recommendation all use t-distributed Stochastic Neighbor Embedding (t-SNE) for visualization.…”

Section: Qualitative Verificationmentioning

confidence: 99%

“…[1] DNA Methylation Data to Predict Suicidal and Non-Suicidal Deaths: A ML. Approach [76] Semi-Supervised and Transfer Learning Approaches for Low Resource Senti. Class.…”

Section: Genericmentioning

confidence: 99%

Specialized document embeddings for aspect-based similarity of research papers

Ostendorff

¹

,

Blume²,

Ruas

³

et al. 2022

Proceedings of the 22nd ACM/IEEE Joint Conference on Digital Libraries

View full text Add to dashboard Cite

Document embeddings and similarity measures underpin contentbased recommender systems, whereby a document is commonly represented as a single generic embedding. However, similarity computed on single vector representations provides only one perspective on document similarity that ignores which aspects make two documents alike. To address this limitation, aspect-based similarity measures have been developed using document segmentation or pairwise multi-class document classification. While segmentation harms the document coherence, the pairwise classification approach scales poorly to large scale corpora. In this paper, we treat aspect-based similarity as a classical vector similarity problem in aspect-specific embedding spaces. We represent a document not as a single generic embedding but as multiple specialized embeddings. Our approach avoids document segmentation and scales linearly w.r.t. the corpus size. In an empirical study, we use the Papers with Code corpus containing 157, 606 research papers and consider the task, method, and dataset of the respective research papers as their aspects. We compare and analyze three generic document embeddings, six specialized document embeddings and a pairwise classification baseline in the context of research paper recommendations. As generic document embeddings, we consider FastText, SciBERT, and SPECTER. To compute the specialized document embeddings, we compare three alternative methods inspired by retrofitting, fine-tuning, and Siamese networks. In our experiments, Siamese SciBERT achieved the highest scores. Additional analyses indicate an implicit bias of the generic document embeddings towards the dataset aspect and against the method aspect of each research paper. Our approach of aspect-based document embeddings mitigates potential risks arising from implicit biases by making them explicit. This can, for example, be used for more diverse and explainable recommendations.

show abstract