1. Introduction

2. Personalization scenarios

1)

Search engine. Search engines are widely employed to provide external knowledge to LLMs, reducing LLMs’ memory burden and alleviating the occurrence of hallucinations in LLMs’ responses.

2)

Recommendation engine. Some works have attempted to alleviate the memory burden of LLMs by equipping them with a recommendation engine as a tool, enabling LLMs to offer recommendations grounded on the item corpus.

2.1. Problem formulation

We consider every data sample as an independent user u, with a set of user records associated with each sample across all tasks. These user records are instrumental in tailoring language models to individual users using their specific data. Consequently, each data sample can be divided into three distinct components: an input sequence, which serves as the model’s input; a target output, which represents the expected model output; and a profile, which encapsulates any supplementary information that can be used to customize the model in accordance with the user’s unique preferences or needs.

A general recommendation problem can be defined as the task of maximizing a utility function M. M serves as a measure of how useful an item s is to a recommendation context c, represented as M: C× S → R, where R is a totally ordered set. Here, the recommendation context c is intended as an information collection composed of the user profile, i.e., the list of her preferred items Suser, and those recommended by the system at recommendation time Srec, such that C = {Suser, Srec}. The primary objective is to select an item s′ ∈ S for each recommendation context c ∈ C that maximizes their utility.

In our case, personalized recommendation in provided via text. For a given textual input t, the goal is to develop a model M that generates personalized output o for user u. This task is formulated as arg max_o p(o|t; u). For each user u, the model M can take advantage of user profile P_u = {(t_u1, o_u1), (t_u2, o_u2), …, (t_um,o_m)} where each (t_ui, o_ui) denotes a pair of input text t and personalized output o of user u (Figure 2).

Each personalization user profile comprises an extensive array of data points specific to that user. However, due to the natural limitations on context length in LLMs and considerations of efficiency and cost, it is feasible to include only a subset of these data points as input prompts. Not all elements within a user’s personalization profile may be directly applicable to the particular task the user intends to undertake. IR plays a selective role in extracting the relevant data from the user profile that includes the current, unseen test case.

To compute personalization for a given sample in personalization user profile (t_ui, o _ui), three steps are used:

(1)

a query construction component that transforms the input t_i into a query q for retrieving from the personalization user’s profile;

(2)

a retrieval component IR(q; P_u) that accepts a query q, a user profile Pu and retrieves entries assumed to be most significant from the user profile; and

(3)

a prompt construction component p that assembles a prompt for user u based on input t_i and the obtained entries.

LLM input is obtained using the following: $t_i^{LLM}={\varphi }_p(t_i,\mathrm{I}\mathrm{R}\left({\varphi }_q\left(t_i\right),Pu\right).{(t}_i^{LLM},o_i)$ is used to train and evaluate LLM.

To express the similarity between a recommending system and an LLM, consider the following. A LLM estimates the probability P(x₁, x₂, ..., x_T ) of observing sentence tokens in the given order, namely estimating the entire sequence’s joint probability. Applying the chain rule, this probability is expressed as

An ideal LLM is expected to autonomously generate text by sequentially selecting individual tokens

Hence both LLM and a recommender learn to predict the most probable next token and next item respectively (Di Palma 2023).

Let us consider an recommendation example. The input is: “I find the purchase history list of user_15466:

4110 -> 4467 -> 4468 -> 4472 I wonder what is the next item to recommend to the user. Can you help me decide?”

The output is: “1581”

(Geng et al. 2022) assigned the user and each item a unique ID. Using a training set with thousands of users (and their purchase histories) and unique items, the LLM is able to learn that certain items are similar to one another and that certain users have inclinations towards certain items (due to the nature of the self-attention mechanism). During the pre-training process across all these purchase sequences, the model essentially goes through a form of collaborative filtering. It sees what users have purchased the same items and what items tend to be purchased together. Combine that with an LLM’s ability to produce contextual embeddings, and we suddenly have a very powerful recommendation system.

In this example, although we do not know what item each ID corresponds to, we can infer that item “1581” was selected due to other users purchasing it along with any of the items that “user_15466” already purchased (Aboufoul 2023)

3. Transformer-based personalization

4. Offline and online functionality

5. LLM Personalization via meta-prompting

6. LLM Personalization via abduction

7. Evaluation

8. Personalized Recommendation in Health

I25.1 - Atherosclerotic heart disease of native coronary artery:

8.4. Neuro-symbolic recommender

Having considered conventional and deep learning–based recommender architectures in health, we now proceed to a hybrid, neuro-symbolic technique. The workflow of the neuro-symbolic recommender is shown in Figure 23.

Its usage includes the following steps:

(1) The physician asks for a recommendation of the most frequently used TKs using the recommender system.

(2) The physician gets a recommendation.

(3) The physician selects the appropriate recommendations and does the treatment based on his own decision.

(4) The new updated information about selected TKs is stored in the database (DB).

The HRS is based on the fuzzy classification system, such that the ICD groups, patients’ age, and sex are used as input parameters to predict the treatment strategy (Figure 24). A DNN method based on a multilayer neural net composed of multiple hidden layers, where every neuron in layer i is fully connected to every other neuron in layer i + 1.

The data contained in electronic health records (EHRs) offer the potential to discover relevant patterns that aim to relate diseases and therapies, and thus discover patterns that could help identify empirical medical guidelines that reflect best practices in the healthcare system. Based on this pattern identification, it is then possible to implement recommendation systems based on the idea that a higher volume of procedures is associated with high-quality models (Ochoa and Mustafa 2021).

Ochoa et al. (2021) implement two AND layers (conjunctions), followed by an OR layer (disjunction), to logically evaluate the nodes Mi, which is a process modeling human reasoning in the decision process. The last layer emulates fuzzy logical operations, based on logical gates, modeled by perceptron with fixed weights and biases, activated by so-called squashing activation functions. These squashing functions approximate the cutting function in the nilpotent logical operators

where β is a real nonzero value that needs to be adjusted to let the model be convergence. Thus, the hidden layers can model a threshold-based nilpotent operator [9, 10]: a conjunction, a disjunction, or even an aggregative operator. This means that the weights of the first layer are to be learned, while the hidden layers of the pre-designed neural block, worked as logical operators with frozen weights and biases.

Ochoa et al. (2021) trained two different models for the final model consumption (Figure 25), one with the whole database (Model 1) and another with a database that selects only positive outcomes (Model 2). The goal of this training method is to make predictions based on the positive outcomes and then evaluate the confidence of the prediction.

The training method for the Siamese recommender system is as follows: Two different models are trained based on the whole data set (Model 1) and a dataset consisting only of positive outcomes (Model 2). After that, both models are used to make different predictions. If there is a matching in the predictions, then Prediction 1 is used as the standard with high confidence; otherwise, predictions from Model 2 are provided but have low confidence

8.5. Graph embedding

Patient Safety Indicators (PSIs), created by the federally managed Agency for Healthcare Research and Quality, aim to assess the frequency of potentially avoidable complications or adverse events that patients encounter while hospitalized. The empirical criteria for measurement are derived from widely accepted parameters, encompassing Inpatient Quality Indicators (IQI), Prevention Quality Indicators (PQI), and Patient Safety Indicators (PSI). These parameters are assessed as aggregated metrics, overlooking the inherent patterns and dynamics in the allocation of therapies and medical procedures to patients.

PSIs specifically measure complications and adverse events from:

(1)

medical conditions after admission

(2)

surgical procedures

(3)

obstetric procedures

PSIs have consistently revealed significant variations in complication and adverse event rates among hospitals. Additionally, there is supporting evidence indicating that elevated rates of complications and adverse events may be linked to deficiencies in the quality of care. There is a broad consensus that healthcare providers can mitigate patient complications and adverse events by enhancing the overall quality of care and safety within the healthcare environment.

The integration of information found in Electronic Health Records (EHRs) and their structured modeling is crucial for enhancing patient care, particularly in the development of personalized and adaptable treatment strategies. There is potential in establishing connections between ICD-10 diagnoses and medical procedures. The use of graph-based representations of health attributes can facilitate both detailed and generalized analyses of IQI.

Utilization of a graph-based data representation is anticipated to be highly effective in clustering “similar” patients. With such a model, patients will be interconnected when there are shared or analogous patterns observed among them. This concept leads to the creation of a network-like structure referred to as a patient graph (Ochoa and Mustafa, 2021). This structure can then be analyzed using Graph Neural Networks (GNN) to discern pertinent labels, specifically suitable medical procedures for recommendation. The successful implementation of this graph framework not only relies on the quality of the underlying diagnostic system but also necessitates a comprehensive understanding of how patients with specific diseases are identified.

Patient clustering is conducted under the assumption that each patient, denoted as 𝑘, who shares the same or similar ICD pattern, is connected to other patients with similar ICDs. This means that patients with akin combinations of morbidities and comorbidities are expected to have similar Treatment Keys (TKs). This approach offers an advantage over traditional clustering methods due to its flexibility, allowing for the easy evaluation of different connectivity patterns in new graphs. Unlike traditional clustering methods like k-means or c-means, this graph-based clustering approach doesn’t necessitate the use of ad-hoc definitions or parameters, as it establishes a direct correlation between ICDs and TKs.

Although embedding methods have demonstrated success in various applications, they have a fundamental limitation: their ability to capture complex patterns is intrinsically constrained by the dimensionality of the embedding space. (Nickel and Kiela, 2017). (Ochoa and Mustafa 2021) perform a patient embedding in a graph object 𝐺, as is shown in Figure 27, such that the corresponding ICD distributions are encoded by an object representing an interlinking between similar patients, an object that can be defined as a similarity score $𝒢$̂_𝑖𝑗, which is essentially the adjacency matrix for the linking between the patient 𝑖 and 𝑗.

The objective is to recommend TKs for a patient depending on her diagnosis (ICDs) and additional metadata, like patients’ age, gender, and pertinence to a specific health center. The TKs and ICDs are represented as a multi-label matrix $ℳ$̂_𝐼𝐶𝐷𝜖ℝ^𝐼×𝑀, where 𝑀 is the number of unique ICDs and 𝐼 is the total number of patients, such that each element in the matrix can be defined as $ℳ$̂_𝐼𝐶𝐷^𝑖𝑚 (Figure 26).

Figure 26. Sample ICD multi-label matrix for the patient population $ℳ$_𝐼𝐶𝐷, where the columns are the ICDs and the rows are the patients.

Figure 26. Sample ICD multi-label matrix for the patient population $ℳ$_𝐼𝐶𝐷, where the columns are the ICDs and the rows are the patients.

Figure 27. Graph encoding of patient’s features.

Figure 27. Graph encoding of patient’s features.

Since a fix clustering method aggregates information in the defined cluster, with graph structures the information is aggregating in each node informing how strong is the similarity between patients, relying on similarity score of (Rocheteau et al. 2021). This similarity is defined by means of a score $𝒢$̂_𝑖𝑗, which is a function that depends on the ICD distribution as well as the gender and age relatedness between patient 𝑖 and patient j

$$d_{ij}={\sum }_{m=1}^M{g}_{ij}^m s_i^m{s}_j^m+c\left[\delta \left({\mu }_i,{\mu }_j\right)\left(1-\left|{\alpha }_i-{\alpha }_j\right|\right)\right]$$

where $g_{ij}^m$ is essentially a weight for the ICD distribution for patients 𝑖 and 𝑗, 𝑚 is the number of elements of each of the 𝑠⃗𝑖 vector, i.e., 𝑠⃗𝑖 = {𝑠_𝑖¹, 𝑠 _𝑖², …, s_𝑖^m, … }, and 𝑀 is the total number of columns of the reduced-dimension $ℳ$̂ ′_𝐼𝐶𝐷 matrix. Additionally, there is a scoring that depends on the patient’s meta parameters, where $\delta \left({\mu }_i,{\mu }_j\right)$ is the Kronecker’s delta (1 if ${\mu }_i={\mu }_j$, 0 otherwise), and 𝑐 is an importance hyper parameter, so that patients with similar parameters get a high score, adjusting the initial ICD scoring in this manner. The goal of this definition is to bias the graph construction to similar patients based not only on similar diagnoses, but also on similar age and gender.

This score is used to define the strength of the linking between two nodes. The corresponding structure of the similarity score $𝒢$̂_𝑖𝑗 is a function that depends on this relatedness 𝑑_𝑖𝑗 and the clustering of similar elements in the graph, and which essentially works as the adjacency matrix linking the nodes 𝑝_𝑖 and 𝑝_𝑗. For the definition of this matrix, neighbor elements should be recognized depending on the distance between 𝑝_𝑖 and 𝑝_𝑗. For this computation we estimate the nearest neighbors using a k-Nearest neighbors’ method by finding a predefined number of training samples closest in distance to the new point and predict the label from that, such that

where 𝑘 is the maximal number of nearest neighbors explored by the clustering algorithm, 𝑎 is an internal calibration parameter of the metric, and 𝑟 is the radius where the clustering is performed.

The graph encoding, as shown in Figure 27, is then used as input for a convolutional neural graph in order to identify corresponding labels for each patient, located in each node of the network. In our case, the labels are the corresponding TKs assigned to each node (patient).

The whole algorithm is presented in Figure 28. One can see a shaped-charge architecture: the DNN is followed by the kNN.

9. Deployment of LLM Personalization

10. Discussion and Conclusions

11. Hands-on

References

1. Abbasian M, Iman Azimi, Amir M. Rahmani, Ramesh Jain (2023) Conversational Health Agents: A Personalized LLM-Powered Agent Framework. https://aps.arxiv.org/abs/2310.02374.
2. Aboufoul M (2023) Using Large Language Models as Recommendation Systems. A review of recent research and a custom implementation. https://towardsdatascience.com/using-large-language-models-as-recommendation-systems-49e8aeeff29b.
3. Bamman D, Chris Dyer, and Noah A. Smith. (2014) Distributed representations of geographically situated language. In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers).
4. Beckers health (2023) https://www.beckershospitalreview.com/healthcare-information-technology/google-receives-more-than-1-billion-health-questions-every-day.html, accessed on 6 February 2021).
5. Bocanegra CL, Sevillano JL, Rizo C, Civit A, Fernandez-Luque L. HealthRecSys: a semantic content-based recommender system to complement health videos. BMC Med Inform Decis Mak. 2017 May 15;17(1):63. [CrossRef]
6. Cai Y, Yu F, Kumar M, Gladney R, Mostafa J. Health Recommender Systems Development, Usage, and Evaluation from 2010 to 2022: A Scoping Review. Int J Environ Res Public Health. 2022 Nov 16;19(22):15115.
7. Carter, E.L.; Nunlee-Bland, G.; Callender, C. 2011. A Patient-Centric, Provider-Assisted Diabetes Telehealth Self-Management Intervention for Urban Minorities. Perspect. Health Inf. Manag.
8. Chen, J., Liu, Z., Huang, X., Wu, C., Liu, Q., Jiang, G., Pu, Y., Lei, Y., Chen, X., Wang, X., Lian, D., & Chen, E. (2023). When Large Language Models Meet Personalization: Perspectives of Challenges and Opportunities. ArXiv, abs/2307.16376.
9. Colab 2023 https://colab.research.google.com/github/Mohammadhia/t5_p5_recommendation_system/blob/main/T5P5_Recommendation_System.ipynb.
10. Dai, W., Xu, Q., Yu, Y., & Zhou, Z. (2019). Bridging Machine Learning and Logical Reasoning by Abductive Learning. Neural Information Processing Systems.
11. De Croon R, Van Houdt L, Htun NN, Štiglic G, Vanden Abeele V, Verbert K. Health Recommender Systems: Systematic Review. J Med Internet Res. 2021 Jun 29;23(6):e18035.
12. Denecker, M. and Kakas, A.C. (2002). “Abduction in Logic Programming”. In Kakas, A.C.; Sadri, F. (eds.). Computational Logic: Logic Programming and Beyond: Essays in Honour of Robert A. Kowalski. Lecture Notes in Computer Science. Vol. 2407. Springer. pp. 402–437.
13. Di Palma D, Giovanni Maria Biancofiore, Vito Walter Anelli, Fedelucio Narducci, Tommaso Di Noia, Eugenio Di Sciascio (2023) Evaluating ChatGPT as a Recommender System: A Rigorous Approach. https://arxiv.org/abs/2309.03613.
14. Ficler, Jessica, and Yoav Goldberg. ”Controlling linguistic style aspects in neural language generation.” arXiv preprint arXiv:1707.02633 (2017).
15. Flek L 2020. Returning the N to NLP: Towards contextually personalized classification models. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 7828–7838. 7828.
16. Friedman L, S. Ahuja, D. Allen, T. Tan, H. Sidahmed, C. Long, J. Xie, G. Schubiner, A. Patel, H. Lara et al., “Leveraging large language models in conversational recommender systems,” arXiv preprint arXiv:2305.07961, 2023.
17. Galitsky B -github (2024) Treatment Recommendation Personalization System. https://github.com/bgalitsky/LLM-personalization/blob/main/Treatment_Recommendation_Personalization_System.ipynb.
18. Galitsky B (2002) Designing the personalized agent for the virtual mental world. AAAI FSS-2002 Symposium on Personalized agents, 21-29.
19. Galitsky B (2005) A Personalized Assistant for Customer Complaints Management Systems.AAAI Spring Symposium: Persistent Assistants: Living and Working with AI, 84-89.
20. Galitsky B (2016) Providing personalized recommendation for attending events based on individual interest profiles. Artif. Intell. Res. 5 (1), 1-13.
21. Galitsky B (2017) Content inversion for user searches and product recommendations systems and methods US Patent 9,336,297.
22. Galitsky B (2021) Recommendation by Joining a Human Conversation. In book: Artificial Intelligence for Customer Relationship Management: Solving Customer Problems, Springer, Cham.
23. Galitsky B, EW McKenna (2017) Sentiment extraction from consumer reviews for providing product recommendations. US Patent 9,646,078.
24. Galitsky B, Kovalerchuk B (2014) Improving web search relevance with learning structure of domain concepts. In: Clusters, Orders, and Trees: Methods and Applications: In Honor of Boris Mirkin’s 70th Birthday. 341-376. Springer New York.
25. Galitsky B(2023) Generating recommendations by using communicative discourse trees of conversations. US Patent 11,599,731. 11,599.
26. Gao T, Adam Fisch, and Danqi Chen. 2021. Making pre-trained language models better few-shot learners. In Proc. of ACL-IJCNLP, pages 3816–3830.
27. Gao Y, T. Sheng, Y. Xiang, Y. Xiong, H. Wang, and J. Zhang, “Chat-rec: Towards interactive and explainable LLMs-augmented recommender system,” arXiv preprint arXiv:2303.14524, 2023. arXiv:2303.14524, 2023.
28. Garimella A, Carmen Banea, and Rada Mihalcea. 2017. Demographic-aware word associations. In nProceedings of the 2017 Conference on Empirical Methods in Natural Language Processing.
29. Geng S, Shuchang Liu, Zuohui Fu, Yingqiang Ge, and Yongfeng Zhang. 2022. Recommendation as Language Processing (RLP): A Unified Pretrain, Personalized Prompt & Predict Paradigm (P5). In Proceedings of the 16th ACM Conference on Recommender Systems (RecSys ‘22). Association for Computing Machinery, New York, NY, USA, 299–315. [CrossRef]
30. Goodman N (2023) Meta-Prompt: A Simple Self-Improving Language Agent. https://noahgoodman.substack.com/p/meta-prompt-a-simple-self-improving.
31. Greylink C (2023) LLMs and meta-prompting. https://cobusgreyling.medium.com/meta-prompt-60d4925b4347. Last downloaded Nov 22, 2023.
32. Gu Y, Xuebing Yang, Lei Tian, Hongyu Yang, Jicheng Lv, Chao Yang, Jinwei Wang, Jianing Xi, Guilan Kong, Wensheng Zhang. 2022 Structure-aware siamese graph neural networks for encounter-level patient similarity learning, Journal of Biomedical Informatics, Volume 127.
33. He, Z., Zhouhang Xie, Rahul Jha, Harald Steck, Dawen Liang, Yesu Feng, Bodhisattwa Prasad Majumder, Nathan Kallus, and Julian McAuley. 2023. Large Language Models as Zero-Shot Conversational Recommenders. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Management (CIKM ’23), October 21–25, 2023, Birmingham, United Kingdom.
34. Hou Y (2023) Awesome LLM-powered agent. https://github.com/hyp1231/awesome-llm-powered-agent.
35. Hou Y, Hongyuan Dong, Xinghao Wang, Bohan Li, and Wanxiang Che. 2022. MetaPrompting: Learning to Learn Better Prompts. In Proceedings of the 29th International Conference on Computational Linguistics, pages 3251–3262.
36. Kakas AC, R. A. Kowalski, and F. Toni. Abductive logic programming. Journal of Logic Computation, 2(6):719–770, 1992.
37. Khaleel I., Wimmer B.C., Peterson G.M., Zaidi S.T.R., Roehrer E., Cummings E., Lee K. Health Information Overload among Health Consumers: A Scoping Review. Patient Educ. Couns. 2020;103:15–32.
38. Kirk HR, Bertie Vidgen, Paul Röttger, and Scott A Hale. 2023. Personalisation within bounds: A risk taxonomy and policy framework for the alignment of large language models with personalised feedback. arXiv preprint arXiv:2303.05453.
39. Lewis, Robert, Ferguson, Craig, Wilks, Chelsey, Jones, Noah and Picard, Rosalind. 2022. Can a Recommender System Support Treatment Personalisation in Digital Mental Health Therapy? A Quantitative Feasibility Assessment Using Data from a Behavioural Activation Therapy App.
40. Li, Xiang Lisa and Percy Liang. 2021. Prefix-tuning: Optimizing continuous prompts for generation. In Proc. of ACL-IJCNLP, pages 4582–4597.
41. Lyu H and Song Jiang and Hanqing Zeng and Qifan Wang and Si Zhang and Ren Chen and Chris Leung and Jiajie Tang and Yinglong Xia and Jiebo Luo (2023) LLM-Rec: Personalized Recommendation via Prompting Large Language Models, arXiv2307.15780.
42. Mork PJ, Bach K; selfBACK Consortium. A Decision Support System to Enhance Self-Management of Low Back Pain: Protocol for the selfBACK Project. JMIR Res Protoc. 2018 Jul 20;7(7):e167.
43. Neisser, U. (1976). Cognition and reality. San Francisco, CA: W. H. Freeman and Company.
44. Nickel, M., and Kiela, D. (2017). Poincar\’e Embeddings for Learning Hierarchical Representations. ArXiv170508039 Cs Stat.
45. Ochoa JGD, Mustafa FE. Graph neural network modelling as a potentially effective method for predicting and analyzing procedures based on patients’ diagnoses. Artif Intell Med. 2022 Sep;131:102359.
46. Ochoa, J.G.D., Csiszár, O. & Schimper, T. Medical recommender systems based on continuous-valued logic and multi-criteria decision operators, using interpretable neural networks. BMC Med Inform Decis Mak 21, 186 (2021).
47. Pew Research Center 2021 https://www.pewresearch.org/internet/2005/05/17/health-information-online/, accessed on 6 February 2021).
48. Rocheteau, E., Tong, C., Veličković, P., Lane, N., and Liò, P. (2021). Predicting Patient Outcomes with Graph Representation Learning. ArXiv210103940 Cs.
49. Sahoo, A.K.; Pradhan, C.; Barik, R.K.; Dubey, H. DeepReco: Deep Learning Based Health Recommender System Using Collaborative Filtering. Computation 2019, 7, 25.
50. Salemi A, S Mysore, M Bendersky, H Zamani (2023) LaMP: When Large Language Models Meet Personalization. arXiv:2304.11406.
51. Skopyk Khrystyna, Artem Chernodub and Vipul Raheja (2022)Personalizing Large Language Models.
52. Sun Y, Zhou J, Ji M, Pei L, Wang Z. Development and Evaluation of Health Recommender Systems: Systematic Scoping Review and Evidence Mapping. J Med Internet Res. 2023.
53. Syed, Bakhtiyar, et al. Adapting language models for non-parallel author-stylized rewriting.” Proceedings of the AAAI Conference on Artificial Intelligence. Vol. 34.
54. No. 05. 2020.”.
55. Wang A, Amanpreet Singh, Julian Michael, Felix Hill, Omer Levy, and Samuel Bowman. 2018. GLUE: A multi-task benchmark and analysis platform for natural language understanding. In Proceedings of the 2018 EMNLP Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP, pages 353–355, Brussels, Belgium. Association for Computational Linguistics.
56. Wei Y, Xiang Wang, Liqiang Nie, Xiangnan He, Richang Hong, and Tat-Seng Chua(2019). MMGCN: Multi-modal Graph Convolution Network for Personalized Recommendation of Micro-video. In ACM MM`19, NICE, France, Oct. 21-25, 2019. 2019; -25.
57. Welch C, Jonathan K. Kummerfeld, Veronica Perez-Rosas, and Rada Mihalcea. 2020. Compositional demographic word embeddings. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing.
58. Welch, Charles & Gu, Chenxi & Kummerfeld, Jonathan & Pérez-Rosas, Verónica & Mihalcea, Rada. (2022). Leveraging Similar Users for Personalized Language Modeling with Limited Data. ACL, 1742-1752.
59. Wiesner M, Pfeifer D. Health recommender systems: concepts, requirements, technical basics and challenges. Int J Environ Res Public Health. 2014 Mar 3;11(3):2580-607.
60. Xue, Gui-Rong & Han, Jie & Yu, Yong & Yang, Qiang (2009) User language model for collaborative personalized search. ACM Trans. Inf. Syst.. 27. [CrossRef]
61. Yang F, Zheng Chen, Ziyan Jiang, Eunah Cho, Xiaojiang Huang, Yanbin Lu (2023) Palr: Personalization aware LLMs for recommendation,” arXiv preprint arXiv:2305.07622.
62. Yang, Diyi, and Lucie Flek. (2021) Towards User-Centric Text-to-Text Generation: A Survey.” International Conference on Text, Speech, and Dialogue. Springer, Cham, 2021.
63. Zhou K, Hui Yu, Wayne Xin Zhao, and Ji-Rong Wen. 2022. Filter-enhanced MLP is All You Need for Sequential Recommendation. In Proceedings of the ACM Web Conference 2022 (WWW ‘22). Association for Computing Machinery, New York, NY, USA, 2388–2399.