A Chatbot with Emojis and Memes

Ling He

doi:10.54097/bf24ap74

Authors

Ling He

DOI:

https://doi.org/10.54097/bf24ap74

Keywords:

Humor; Emoji; Meme; Llama2 fine-tune.

Abstract

With the consistent development of information technology, online social life has become increasingly diverse and amusing. In the current online communication environment, emojis and memes have become common media of interaction. Compared to traditional text-based communication, emojis and memes not only add a sense of warmth and humor to conversations but also present new opportunities and challenges in the field of large language models (LLMs). This paper introduces a chatbot designed to enhance user interaction, consisting of three main modules: a dialogue generation module, an emoji embedding module, and a meme generation module. This study first fine-tunes the Llama2-7b-chat-hf pretrained model using the OogiriGo dataset to build a more humorous language style. Then, it incorporates EmojiLM and Supermeme individually for emoji embedding and meme generation. The results section of this paper presents the details of the evolution of the chatbot’s language style. Finally, the paper discusses potential improvements to the model and explores the future development prospects and challenges in related fields.

Downloads

Download data is not yet available.

References

[1] Liao Y, Wang Y, Liu Q, & Jiang X. Gpt-based generation for classical chinese poetry. arXiv preprint arXiv:1907.00151. 2019.

[2] Barbieri F, Anke L E, Camacho-Collados J, Schockaert S, & Saggion H. Interpretable emoji prediction via label-wise attention LSTMs. In Proceedings of the 2018 conference on empirical methods in natural language processing 2018, (pp. 4766-4771).

[3] Zhao S, Jiang H, Tao H, Zha R, Zhang K, Xu T, & Chen E. Pedm: A multi-task learning model for persona-aware emoji-embedded dialogue generation. ACM Transactions on Multimedia Computing, Communications and Applications, 2023, 19(3s), 1-21.

[4] Kusal S, Patil S, Choudrie J, Kotecha K, Mishra S, & Abraham A. AI-based conversational agents: a scoping review from technologies to future directions. IEEE Access, 2022, 10, 92337-92356.

[5] Ainslie J, Lee-Thorp J, de Jong M, Zemlyanskiy Y, Lebrón F, & Sanghai S. Gqa: Training generalized multi-query transformer models from multi-head checkpoints. arXiv preprint arXiv:2305.13245. 2023.

[6] Touvron H, Martin L, Stone K, Albert P, Almahairi A, Babaei Y, ... & Scialom T. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288. 2023.

[7] Zhong S, Huang Z, Gao S, Wen W, Lin L, Zitnik M, & Zhou P. Let's Think Outside the Box: Exploring Leap-of-Thought in Large Language Models with Creative Humor Generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2024, (pp. 13246-13257).

[8] Zheng Y, Zhang R, Zhang J, Ye Y, Luo Z, Feng Z, & Ma Y. Llamafactory: Unified efficient fine-tuning of 100+ language models. arXiv preprint arXiv:2403.13372. 2024.

[9] Peng L, Wang Z, Liu H, Wang Z, & Shang J. EmojiLM: Modeling the New Emoji Language. arXiv preprint arXiv:2311.01751. 2023.

[10] Eisner B, Rocktäschel T, Augenstein I, Bošnjak M, & Riedel S. emoji2vec: Learning emoji representations from their description. arXiv preprint arXiv:1609.08359. 2016.

[11] Lee S, Jeong D, & Park E. MultiEmo: Multi-task framework for emoji prediction. Knowledge-Based Systems, 2022, 242, 108437.

[12] Peirson V A L, & Tolunay E M. Dank learning: Generating memes using deep neural networks. arXiv preprint arXiv:1806.04510. 2018.

[13] Ranjan A, Srivastava V, Khatri J, Bhat S, & Karande S. Meme Generation with Multi-modal Input and Planning. In Proceedings of the 2nd International Workshop on Deep Multimodal Generation and Retrieval 2024, (pp. 21-25).

[14] Agarwal S, Sharma S, Nakov P, & Chakraborty T. MemeMQA: Multimodal Question Answering for Memes via Rationale-Based Inferencing. arXiv preprint arXiv:2405.11215. 2024.

[15] Yntec (n.d.). Memento. Retrieved December 19, 2024, from https://huggingface.co/Yntec/Memento.