ID-Agnostic User Behavior Pre-training for Sequential Recommendation

Recently, sequential recommendation has emerged as a widely studied topic. Existing researches mainly design effective neural architectures to model user behavior sequences based on item IDs. However, this kind of approach highly relies on user-item interaction data and neglects the attribute- or ch...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Mu, Shanlei, Hou, Yupeng, Zhao, Wayne Xin, Li, Yaliang, Ding, Bolin
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Mu, Shanlei
Hou, Yupeng
Zhao, Wayne Xin
Li, Yaliang
Ding, Bolin
description Recently, sequential recommendation has emerged as a widely studied topic. Existing researches mainly design effective neural architectures to model user behavior sequences based on item IDs. However, this kind of approach highly relies on user-item interaction data and neglects the attribute- or characteristic-level correlations among similar items preferred by a user. In light of these issues, we propose IDA-SR, which stands for ID-Agnostic User Behavior Pre-training approach for Sequential Recommendation. Instead of explicitly learning representations for item IDs, IDA-SR directly learns item representations from rich text information. To bridge the gap between text semantics and sequential user behaviors, we utilize the pre-trained language model as text encoder, and conduct a pre-training architecture on the sequential user behaviors. In this way, item text can be directly utilized for sequential recommendation without relying on item IDs. Extensive experiments show that the proposed approach can achieve comparable results when only using ID-agnostic item representations, and performs better than baselines by a large margin when fine-tuned with ID information.
doi_str_mv 10.48550/arxiv.2206.02323
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2206_02323</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2206_02323</sourcerecordid><originalsourceid>FETCH-LOGICAL-a673-16d3ae3a2806033e1f228d1ed9fc5f6f76436c48ff188beee978077d5c44f663</originalsourceid><addsrcrecordid>eNotz71OwzAUhmEvDKhwAUz1DTg4Po7tjqXlp1IlEC1zdLCPi6XGASdUcPfQwvRJ7_BJD2NXtay0axp5jeUrHSqlpKmkAgXnbLlaivku98OYPH8ZqPAbesND6gt_KiTGgimnvOPxN2zo45PymHDPn8n3XUc54Jj6fMHOIu4HuvzfCdvc3W4XD2L9eL9azNcCjQVRmwBIgMpJIwGojkq5UFOYRd9EE63RYLx2MdbOvRLRzDppbWi81tEYmLDp3-tJ0b6X1GH5bo-a9qSBH9haRLc</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>ID-Agnostic User Behavior Pre-training for Sequential Recommendation</title><source>arXiv.org</source><creator>Mu, Shanlei ; Hou, Yupeng ; Zhao, Wayne Xin ; Li, Yaliang ; Ding, Bolin</creator><creatorcontrib>Mu, Shanlei ; Hou, Yupeng ; Zhao, Wayne Xin ; Li, Yaliang ; Ding, Bolin</creatorcontrib><description>Recently, sequential recommendation has emerged as a widely studied topic. Existing researches mainly design effective neural architectures to model user behavior sequences based on item IDs. However, this kind of approach highly relies on user-item interaction data and neglects the attribute- or characteristic-level correlations among similar items preferred by a user. In light of these issues, we propose IDA-SR, which stands for ID-Agnostic User Behavior Pre-training approach for Sequential Recommendation. Instead of explicitly learning representations for item IDs, IDA-SR directly learns item representations from rich text information. To bridge the gap between text semantics and sequential user behaviors, we utilize the pre-trained language model as text encoder, and conduct a pre-training architecture on the sequential user behaviors. In this way, item text can be directly utilized for sequential recommendation without relying on item IDs. Extensive experiments show that the proposed approach can achieve comparable results when only using ID-agnostic item representations, and performs better than baselines by a large margin when fine-tuned with ID information.</description><identifier>DOI: 10.48550/arxiv.2206.02323</identifier><language>eng</language><subject>Computer Science - Information Retrieval</subject><creationdate>2022-06</creationdate><rights>http://creativecommons.org/licenses/by/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2206.02323$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2206.02323$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Mu, Shanlei</creatorcontrib><creatorcontrib>Hou, Yupeng</creatorcontrib><creatorcontrib>Zhao, Wayne Xin</creatorcontrib><creatorcontrib>Li, Yaliang</creatorcontrib><creatorcontrib>Ding, Bolin</creatorcontrib><title>ID-Agnostic User Behavior Pre-training for Sequential Recommendation</title><description>Recently, sequential recommendation has emerged as a widely studied topic. Existing researches mainly design effective neural architectures to model user behavior sequences based on item IDs. However, this kind of approach highly relies on user-item interaction data and neglects the attribute- or characteristic-level correlations among similar items preferred by a user. In light of these issues, we propose IDA-SR, which stands for ID-Agnostic User Behavior Pre-training approach for Sequential Recommendation. Instead of explicitly learning representations for item IDs, IDA-SR directly learns item representations from rich text information. To bridge the gap between text semantics and sequential user behaviors, we utilize the pre-trained language model as text encoder, and conduct a pre-training architecture on the sequential user behaviors. In this way, item text can be directly utilized for sequential recommendation without relying on item IDs. Extensive experiments show that the proposed approach can achieve comparable results when only using ID-agnostic item representations, and performs better than baselines by a large margin when fine-tuned with ID information.</description><subject>Computer Science - Information Retrieval</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2022</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz71OwzAUhmEvDKhwAUz1DTg4Po7tjqXlp1IlEC1zdLCPi6XGASdUcPfQwvRJ7_BJD2NXtay0axp5jeUrHSqlpKmkAgXnbLlaivku98OYPH8ZqPAbesND6gt_KiTGgimnvOPxN2zo45PymHDPn8n3XUc54Jj6fMHOIu4HuvzfCdvc3W4XD2L9eL9azNcCjQVRmwBIgMpJIwGojkq5UFOYRd9EE63RYLx2MdbOvRLRzDppbWi81tEYmLDp3-tJ0b6X1GH5bo-a9qSBH9haRLc</recordid><startdate>20220605</startdate><enddate>20220605</enddate><creator>Mu, Shanlei</creator><creator>Hou, Yupeng</creator><creator>Zhao, Wayne Xin</creator><creator>Li, Yaliang</creator><creator>Ding, Bolin</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20220605</creationdate><title>ID-Agnostic User Behavior Pre-training for Sequential Recommendation</title><author>Mu, Shanlei ; Hou, Yupeng ; Zhao, Wayne Xin ; Li, Yaliang ; Ding, Bolin</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a673-16d3ae3a2806033e1f228d1ed9fc5f6f76436c48ff188beee978077d5c44f663</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2022</creationdate><topic>Computer Science - Information Retrieval</topic><toplevel>online_resources</toplevel><creatorcontrib>Mu, Shanlei</creatorcontrib><creatorcontrib>Hou, Yupeng</creatorcontrib><creatorcontrib>Zhao, Wayne Xin</creatorcontrib><creatorcontrib>Li, Yaliang</creatorcontrib><creatorcontrib>Ding, Bolin</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Mu, Shanlei</au><au>Hou, Yupeng</au><au>Zhao, Wayne Xin</au><au>Li, Yaliang</au><au>Ding, Bolin</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>ID-Agnostic User Behavior Pre-training for Sequential Recommendation</atitle><date>2022-06-05</date><risdate>2022</risdate><abstract>Recently, sequential recommendation has emerged as a widely studied topic. Existing researches mainly design effective neural architectures to model user behavior sequences based on item IDs. However, this kind of approach highly relies on user-item interaction data and neglects the attribute- or characteristic-level correlations among similar items preferred by a user. In light of these issues, we propose IDA-SR, which stands for ID-Agnostic User Behavior Pre-training approach for Sequential Recommendation. Instead of explicitly learning representations for item IDs, IDA-SR directly learns item representations from rich text information. To bridge the gap between text semantics and sequential user behaviors, we utilize the pre-trained language model as text encoder, and conduct a pre-training architecture on the sequential user behaviors. In this way, item text can be directly utilized for sequential recommendation without relying on item IDs. Extensive experiments show that the proposed approach can achieve comparable results when only using ID-agnostic item representations, and performs better than baselines by a large margin when fine-tuned with ID information.</abstract><doi>10.48550/arxiv.2206.02323</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2206.02323
ispartof
issn
language eng
recordid cdi_arxiv_primary_2206_02323
source arXiv.org
subjects Computer Science - Information Retrieval
title ID-Agnostic User Behavior Pre-training for Sequential Recommendation
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-19T18%3A49%3A15IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=ID-Agnostic%20User%20Behavior%20Pre-training%20for%20Sequential%20Recommendation&rft.au=Mu,%20Shanlei&rft.date=2022-06-05&rft_id=info:doi/10.48550/arxiv.2206.02323&rft_dat=%3Carxiv_GOX%3E2206_02323%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true