paper-with-me

홈 › Papers

FeelsGoodMan: Inferring Semantics of Twitch Neologisms

2021-08-18 · Pavel Dolin, Luc d'Hauthuille, Andrea Vattani

Twitch chats pose a unique problem in natural language understanding due to a large presence of neologisms, specifically emotes. There are a total of 8.06 million emotes, over 400k of which were used in the week studied. There is virtually no information on the meaning or sentiment of emotes, and with a constant influx of new emotes and drift in their frequencies, it becomes impossible to maintain an updated manually-labeled dataset. Our paper makes a two fold contribution. First we establish a new baseline for sentiment analysis on Twitch data, outperforming the previous supervised benchmark by 7.9% points. Secondly, we introduce a simple but powerful unsupervised framework based on word embeddings and k-NN to enrich existing models with out-of-vocabulary knowledge. This framework allows us to auto-generate a pseudo-dictionary of emotes and we show that we can nearly match the supervised benchmark above even when injecting such emote knowledge into sentiment classifiers trained on extraneous datasets such as movie reviews or Twitter.

📄 PDF Abstract BibTeX arXiv:2108.08411

Code (0)

등록된 구현이 없습니다.

Tasks

Natural Language UnderstandingSentiment AnalysisWord Embeddings

Methods 이 논문이 사용한 방법론

k-NN $k$-Nearest Neighbors is a clustering-based algorithm for classification and regression. It is a a type of instance-based learning as it does not attempt to construct a…

Similar Papers 제목 키워드 기반

FeelsGoodMan: Inferring Semantics of Twitch Neologisms

2021-11-16 · ACL ARR November 2021 11 · Anonymous

Twitch chat messages pose a unique problem in natural language understanding due to a large presence of neologisms, specifically emotes. There are a total of 8.06 million emotes, over 400k of which were observed during t…

Natural Language UnderstandingSentiment AnalysisWord Embeddings

NEO-BENCH: Evaluating Robustness of Large Language Models with Neologisms

2024-02-19 · Jonathan Zheng, Alan Ritter, Wei Xu

The performance of Large Language Models (LLMs) degrades from the temporal drift between data used for model training and newer text seen during inference. One understudied avenue of language change causing data drift is…

Machine TranslationNatural Language UnderstandingSentence

ToxiTwitch: Toward Emote-Aware Hybrid Moderation for Live Streaming Platforms

2026-01-22 · Baktash Ansari, Elias Martin, Afra Mashhadi arxiv

The rapid growth of live-streaming platforms such as Twitch has introduced complex challenges in moderating toxic behavior. Traditional moderation approaches, such as human annotation and keyword-based filtering, have de…

A primer on getting neologisms from foreign languages to under-resourced languages

2023-03-07 · Luis Camacho

Mainly due to lack of support, most under-resourced languages have a reduced lexicon in most realms and domains of increasing importance, then their speakers need to significantly augment it. Although neologisms should a…

Common Sense Reasoning

Unsupervised Neologism Normalization Using Embedding Space Mapping

2019-11-01 · WS 2019 11 · Nasser Zalmout, Kapil Thadani, Aasish Pappu

This paper presents an approach for detecting and normalizing neologisms in social media content. Neologisms refer to recent expressions that are specific to certain entities or events and are being increasingly used by …

Natural Language UnderstandingText Normalization