Text Matching Improves Sequential Recommendation by Reducing Popularity Biases

This paper proposes Text mAtching based SequenTial rEcommendation model (TASTE), which maps items and users in an embedding space and recommends items by matching their text representations. TASTE verbalizes items and user-item interactions using identifiers and attributes of items. To better charac...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Liu, Zhenghao, Mei, Sen, Xiong, Chenyan, Li, Xiaohua, Yu, Shi, Liu, Zhiyuan, Gu, Yu, Yu, Ge
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 Liu, Zhenghao
Mei, Sen
Xiong, Chenyan
Li, Xiaohua
Yu, Shi
Liu, Zhiyuan
Gu, Yu
Yu, Ge
description This paper proposes Text mAtching based SequenTial rEcommendation model (TASTE), which maps items and users in an embedding space and recommends items by matching their text representations. TASTE verbalizes items and user-item interactions using identifiers and attributes of items. To better characterize user behaviors, TASTE additionally proposes an attention sparsity method, which enables TASTE to model longer user-item interactions by reducing the self-attention computations during encoding. Our experiments show that TASTE outperforms the state-of-the-art methods on widely used sequential recommendation datasets. TASTE alleviates the cold start problem by representing long-tail items using full-text modeling and bringing the benefits of pretrained language models to recommendation systems. Our further analyses illustrate that TASTE significantly improves the recommendation accuracy by reducing the popularity bias of previous item id based recommendation models and returning more appropriate and text-relevant items to satisfy users. All codes are available at https://github.com/OpenMatch/TASTE.
doi_str_mv 10.48550/arxiv.2308.14029
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2308_14029</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2308_14029</sourcerecordid><originalsourceid>FETCH-LOGICAL-a679-3b295317ac524390c5a2884761b5ba9675ebe03acba3733ab9fb578fe8c2ce6e3</originalsourceid><addsrcrecordid>eNotz8tOwzAUBFBvWKDCB7DCP5Dg-MaxvYSKR6XyEGQfXbs3YCkvEqdq_h5aWI00Go10GLvKRJobpcQNjoewTyUIk2a5kPacvZR0iPwZo_8K3SfftMPY72niH_Q9UxcDNvydfN-21O0whr7jbvltdrM_zt_6YW5wDHHhdwEnmi7YWY3NRJf_uWLlw325fkq2r4-b9e02wULbBJy0CjKNXskcrPAKpTG5LjKnHNpCK3IkAL1D0ADobO2UNjUZLz0VBCt2_Xd7AlXDGFocl-oIq04w-AHo0UmZ</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Text Matching Improves Sequential Recommendation by Reducing Popularity Biases</title><source>arXiv.org</source><creator>Liu, Zhenghao ; Mei, Sen ; Xiong, Chenyan ; Li, Xiaohua ; Yu, Shi ; Liu, Zhiyuan ; Gu, Yu ; Yu, Ge</creator><creatorcontrib>Liu, Zhenghao ; Mei, Sen ; Xiong, Chenyan ; Li, Xiaohua ; Yu, Shi ; Liu, Zhiyuan ; Gu, Yu ; Yu, Ge</creatorcontrib><description>This paper proposes Text mAtching based SequenTial rEcommendation model (TASTE), which maps items and users in an embedding space and recommends items by matching their text representations. TASTE verbalizes items and user-item interactions using identifiers and attributes of items. To better characterize user behaviors, TASTE additionally proposes an attention sparsity method, which enables TASTE to model longer user-item interactions by reducing the self-attention computations during encoding. Our experiments show that TASTE outperforms the state-of-the-art methods on widely used sequential recommendation datasets. TASTE alleviates the cold start problem by representing long-tail items using full-text modeling and bringing the benefits of pretrained language models to recommendation systems. Our further analyses illustrate that TASTE significantly improves the recommendation accuracy by reducing the popularity bias of previous item id based recommendation models and returning more appropriate and text-relevant items to satisfy users. All codes are available at https://github.com/OpenMatch/TASTE.</description><identifier>DOI: 10.48550/arxiv.2308.14029</identifier><language>eng</language><subject>Computer Science - Artificial Intelligence ; Computer Science - Information Retrieval</subject><creationdate>2023-08</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>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,781,886</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2308.14029$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2308.14029$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Liu, Zhenghao</creatorcontrib><creatorcontrib>Mei, Sen</creatorcontrib><creatorcontrib>Xiong, Chenyan</creatorcontrib><creatorcontrib>Li, Xiaohua</creatorcontrib><creatorcontrib>Yu, Shi</creatorcontrib><creatorcontrib>Liu, Zhiyuan</creatorcontrib><creatorcontrib>Gu, Yu</creatorcontrib><creatorcontrib>Yu, Ge</creatorcontrib><title>Text Matching Improves Sequential Recommendation by Reducing Popularity Biases</title><description>This paper proposes Text mAtching based SequenTial rEcommendation model (TASTE), which maps items and users in an embedding space and recommends items by matching their text representations. TASTE verbalizes items and user-item interactions using identifiers and attributes of items. To better characterize user behaviors, TASTE additionally proposes an attention sparsity method, which enables TASTE to model longer user-item interactions by reducing the self-attention computations during encoding. Our experiments show that TASTE outperforms the state-of-the-art methods on widely used sequential recommendation datasets. TASTE alleviates the cold start problem by representing long-tail items using full-text modeling and bringing the benefits of pretrained language models to recommendation systems. Our further analyses illustrate that TASTE significantly improves the recommendation accuracy by reducing the popularity bias of previous item id based recommendation models and returning more appropriate and text-relevant items to satisfy users. All codes are available at https://github.com/OpenMatch/TASTE.</description><subject>Computer Science - Artificial Intelligence</subject><subject>Computer Science - Information Retrieval</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotz8tOwzAUBFBvWKDCB7DCP5Dg-MaxvYSKR6XyEGQfXbs3YCkvEqdq_h5aWI00Go10GLvKRJobpcQNjoewTyUIk2a5kPacvZR0iPwZo_8K3SfftMPY72niH_Q9UxcDNvydfN-21O0whr7jbvltdrM_zt_6YW5wDHHhdwEnmi7YWY3NRJf_uWLlw325fkq2r4-b9e02wULbBJy0CjKNXskcrPAKpTG5LjKnHNpCK3IkAL1D0ADobO2UNjUZLz0VBCt2_Xd7AlXDGFocl-oIq04w-AHo0UmZ</recordid><startdate>20230827</startdate><enddate>20230827</enddate><creator>Liu, Zhenghao</creator><creator>Mei, Sen</creator><creator>Xiong, Chenyan</creator><creator>Li, Xiaohua</creator><creator>Yu, Shi</creator><creator>Liu, Zhiyuan</creator><creator>Gu, Yu</creator><creator>Yu, Ge</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20230827</creationdate><title>Text Matching Improves Sequential Recommendation by Reducing Popularity Biases</title><author>Liu, Zhenghao ; Mei, Sen ; Xiong, Chenyan ; Li, Xiaohua ; Yu, Shi ; Liu, Zhiyuan ; Gu, Yu ; Yu, Ge</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a679-3b295317ac524390c5a2884761b5ba9675ebe03acba3733ab9fb578fe8c2ce6e3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Computer Science - Artificial Intelligence</topic><topic>Computer Science - Information Retrieval</topic><toplevel>online_resources</toplevel><creatorcontrib>Liu, Zhenghao</creatorcontrib><creatorcontrib>Mei, Sen</creatorcontrib><creatorcontrib>Xiong, Chenyan</creatorcontrib><creatorcontrib>Li, Xiaohua</creatorcontrib><creatorcontrib>Yu, Shi</creatorcontrib><creatorcontrib>Liu, Zhiyuan</creatorcontrib><creatorcontrib>Gu, Yu</creatorcontrib><creatorcontrib>Yu, Ge</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Liu, Zhenghao</au><au>Mei, Sen</au><au>Xiong, Chenyan</au><au>Li, Xiaohua</au><au>Yu, Shi</au><au>Liu, Zhiyuan</au><au>Gu, Yu</au><au>Yu, Ge</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Text Matching Improves Sequential Recommendation by Reducing Popularity Biases</atitle><date>2023-08-27</date><risdate>2023</risdate><abstract>This paper proposes Text mAtching based SequenTial rEcommendation model (TASTE), which maps items and users in an embedding space and recommends items by matching their text representations. TASTE verbalizes items and user-item interactions using identifiers and attributes of items. To better characterize user behaviors, TASTE additionally proposes an attention sparsity method, which enables TASTE to model longer user-item interactions by reducing the self-attention computations during encoding. Our experiments show that TASTE outperforms the state-of-the-art methods on widely used sequential recommendation datasets. TASTE alleviates the cold start problem by representing long-tail items using full-text modeling and bringing the benefits of pretrained language models to recommendation systems. Our further analyses illustrate that TASTE significantly improves the recommendation accuracy by reducing the popularity bias of previous item id based recommendation models and returning more appropriate and text-relevant items to satisfy users. All codes are available at https://github.com/OpenMatch/TASTE.</abstract><doi>10.48550/arxiv.2308.14029</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2308.14029
ispartof
issn
language eng
recordid cdi_arxiv_primary_2308_14029
source arXiv.org
subjects Computer Science - Artificial Intelligence
Computer Science - Information Retrieval
title Text Matching Improves Sequential Recommendation by Reducing Popularity Biases
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-16T05%3A37%3A12IST&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=Text%20Matching%20Improves%20Sequential%20Recommendation%20by%20Reducing%20Popularity%20Biases&rft.au=Liu,%20Zhenghao&rft.date=2023-08-27&rft_id=info:doi/10.48550/arxiv.2308.14029&rft_dat=%3Carxiv_GOX%3E2308_14029%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