Extracting Implicit User Preferences in Conversational Recommender Systems Using Large Language Models
Conversational recommender systems (CRSs) have garnered increasing attention for their ability to provide personalized recommendations through natural language interactions. Although large language models (LLMs) have shown potential in recommendation systems owing to their superior language understa...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/13/2/221 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832588076026691584 |
|---|---|
| author | Woo-Seok Kim Seongho Lim Gun-Woo Kim Sang-Min Choi |
| author_facet | Woo-Seok Kim Seongho Lim Gun-Woo Kim Sang-Min Choi |
| author_sort | Woo-Seok Kim |
| collection | DOAJ |
| description | Conversational recommender systems (CRSs) have garnered increasing attention for their ability to provide personalized recommendations through natural language interactions. Although large language models (LLMs) have shown potential in recommendation systems owing to their superior language understanding and reasoning capabilities, extracting and utilizing implicit user preferences from conversations remains a formidable challenge. This paper proposes a method that leverages LLMs to extract implicit preferences and explicitly incorporate them into the recommendation process. Initially, LLMs identify implicit user preferences from conversations, which are then refined into fine-grained numerical values using a BERT-based multi-label classifier to enhance recommendation precision. The proposed approach is validated through experiments on three comprehensive datasets: the Reddit Movie Dataset (8413 dialogues), Inspired (825 dialogues), and ReDial (2311 dialogues). Results show that our approach considerably outperforms traditional CRS methods, achieving a 23.3% improvement in Recall@20 on the ReDial dataset and a 7.2% average improvement in recommendation accuracy across all datasets with GPT-3.5-turbo and GPT-4. These findings highlight the potential of using LLMs to extract and utilize implicit conversational information, effectively enhancing the quality of recommendations in CRSs. |
| format | Article |
| id | doaj-art-988f43a7dbbb45d981a37e6fb2227bde |
| institution | Kabale University |
| issn | 2227-7390 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj-art-988f43a7dbbb45d981a37e6fb2227bde2025-01-24T13:39:47ZengMDPI AGMathematics2227-73902025-01-0113222110.3390/math13020221Extracting Implicit User Preferences in Conversational Recommender Systems Using Large Language ModelsWoo-Seok Kim0Seongho Lim1Gun-Woo Kim2Sang-Min Choi3Department of Computer Science and Engineering, Gyeongsang National University, Jinju 52828, Republic of KoreaDigital Division, National Forensic Service, Wonju 26460, Republic of KoreaDepartment of Computer Science and Engineering, Gyeongsang National University, Jinju 52828, Republic of KoreaDepartment of Computer Science and Engineering, Gyeongsang National University, Jinju 52828, Republic of KoreaConversational recommender systems (CRSs) have garnered increasing attention for their ability to provide personalized recommendations through natural language interactions. Although large language models (LLMs) have shown potential in recommendation systems owing to their superior language understanding and reasoning capabilities, extracting and utilizing implicit user preferences from conversations remains a formidable challenge. This paper proposes a method that leverages LLMs to extract implicit preferences and explicitly incorporate them into the recommendation process. Initially, LLMs identify implicit user preferences from conversations, which are then refined into fine-grained numerical values using a BERT-based multi-label classifier to enhance recommendation precision. The proposed approach is validated through experiments on three comprehensive datasets: the Reddit Movie Dataset (8413 dialogues), Inspired (825 dialogues), and ReDial (2311 dialogues). Results show that our approach considerably outperforms traditional CRS methods, achieving a 23.3% improvement in Recall@20 on the ReDial dataset and a 7.2% average improvement in recommendation accuracy across all datasets with GPT-3.5-turbo and GPT-4. These findings highlight the potential of using LLMs to extract and utilize implicit conversational information, effectively enhancing the quality of recommendations in CRSs.https://www.mdpi.com/2227-7390/13/2/221conversational recommender systemslarge language modelsimplicit user preferenceclassification |
| spellingShingle | Woo-Seok Kim Seongho Lim Gun-Woo Kim Sang-Min Choi Extracting Implicit User Preferences in Conversational Recommender Systems Using Large Language Models Mathematics conversational recommender systems large language models implicit user preference classification |
| title | Extracting Implicit User Preferences in Conversational Recommender Systems Using Large Language Models |
| title_full | Extracting Implicit User Preferences in Conversational Recommender Systems Using Large Language Models |
| title_fullStr | Extracting Implicit User Preferences in Conversational Recommender Systems Using Large Language Models |
| title_full_unstemmed | Extracting Implicit User Preferences in Conversational Recommender Systems Using Large Language Models |
| title_short | Extracting Implicit User Preferences in Conversational Recommender Systems Using Large Language Models |
| title_sort | extracting implicit user preferences in conversational recommender systems using large language models |
| topic | conversational recommender systems large language models implicit user preference classification |
| url | https://www.mdpi.com/2227-7390/13/2/221 |
| work_keys_str_mv | AT wooseokkim extractingimplicituserpreferencesinconversationalrecommendersystemsusinglargelanguagemodels AT seongholim extractingimplicituserpreferencesinconversationalrecommendersystemsusinglargelanguagemodels AT gunwookim extractingimplicituserpreferencesinconversationalrecommendersystemsusinglargelanguagemodels AT sangminchoi extractingimplicituserpreferencesinconversationalrecommendersystemsusinglargelanguagemodels |