Memory Augmented Cross-encoders for Controllable Personalized Search

Personalized search represents a problem where retrieval models condition on historical user interaction data in order to improve retrieval results. However, personalization is commonly perceived as opaque and not amenable to control by users. Further, personalization necessarily limits the space of...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Mysore, Sheshera, Dhanania, Garima, Patil, Kishor, Kallumadi, Surya, McCallum, Andrew, Zamani, Hamed
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 Mysore, Sheshera
Dhanania, Garima
Patil, Kishor
Kallumadi, Surya
McCallum, Andrew
Zamani, Hamed
description Personalized search represents a problem where retrieval models condition on historical user interaction data in order to improve retrieval results. However, personalization is commonly perceived as opaque and not amenable to control by users. Further, personalization necessarily limits the space of items that users are exposed to. Therefore, prior work notes a tension between personalization and users' ability for discovering novel items. While discovery of novel items in personalization setups may be resolved through search result diversification, these approaches do little to allow user control over personalization. Therefore, in this paper, we introduce an approach for controllable personalized search. Our model, CtrlCE presents a novel cross-encoder model augmented with an editable memory constructed from users historical items. Our proposed memory augmentation allows cross-encoder models to condition on large amounts of historical user data and supports interaction from users permitting control over personalization. Further, controllable personalization for search must account for queries which don't require personalization, and in turn user control. For this, we introduce a calibrated mixing model which determines when personalization is necessary. This allows system designers using CtrlCE to only obtain user input for control when necessary. In multiple datasets of personalized search, we show CtrlCE to result in effective personalization as well as fulfill various key goals for controllable personalized search.
doi_str_mv 10.48550/arxiv.2411.02790
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2411_02790</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2411_02790</sourcerecordid><originalsourceid>FETCH-arxiv_primary_2411_027903</originalsourceid><addsrcrecordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjE01DMwMrc04GRw8U3NzS-qVHAsTc9NzStJTVFwLsovLtZNzUvOT0ktKlZIyy9ScM7PKynKz8lJTMpJVQgAiubnJeZkVgEVB6cmFiVn8DCwpiXmFKfyQmluBnk31xBnD12wffEFRZm5iUWV8SB748H2GhNWAQAcMziR</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Memory Augmented Cross-encoders for Controllable Personalized Search</title><source>arXiv.org</source><creator>Mysore, Sheshera ; Dhanania, Garima ; Patil, Kishor ; Kallumadi, Surya ; McCallum, Andrew ; Zamani, Hamed</creator><creatorcontrib>Mysore, Sheshera ; Dhanania, Garima ; Patil, Kishor ; Kallumadi, Surya ; McCallum, Andrew ; Zamani, Hamed</creatorcontrib><description>Personalized search represents a problem where retrieval models condition on historical user interaction data in order to improve retrieval results. However, personalization is commonly perceived as opaque and not amenable to control by users. Further, personalization necessarily limits the space of items that users are exposed to. Therefore, prior work notes a tension between personalization and users' ability for discovering novel items. While discovery of novel items in personalization setups may be resolved through search result diversification, these approaches do little to allow user control over personalization. Therefore, in this paper, we introduce an approach for controllable personalized search. Our model, CtrlCE presents a novel cross-encoder model augmented with an editable memory constructed from users historical items. Our proposed memory augmentation allows cross-encoder models to condition on large amounts of historical user data and supports interaction from users permitting control over personalization. Further, controllable personalization for search must account for queries which don't require personalization, and in turn user control. For this, we introduce a calibrated mixing model which determines when personalization is necessary. This allows system designers using CtrlCE to only obtain user input for control when necessary. In multiple datasets of personalized search, we show CtrlCE to result in effective personalization as well as fulfill various key goals for controllable personalized search.</description><identifier>DOI: 10.48550/arxiv.2411.02790</identifier><language>eng</language><subject>Computer Science - Computation and Language ; Computer Science - Information Retrieval</subject><creationdate>2024-11</creationdate><rights>http://creativecommons.org/licenses/by-sa/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/2411.02790$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2411.02790$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Mysore, Sheshera</creatorcontrib><creatorcontrib>Dhanania, Garima</creatorcontrib><creatorcontrib>Patil, Kishor</creatorcontrib><creatorcontrib>Kallumadi, Surya</creatorcontrib><creatorcontrib>McCallum, Andrew</creatorcontrib><creatorcontrib>Zamani, Hamed</creatorcontrib><title>Memory Augmented Cross-encoders for Controllable Personalized Search</title><description>Personalized search represents a problem where retrieval models condition on historical user interaction data in order to improve retrieval results. However, personalization is commonly perceived as opaque and not amenable to control by users. Further, personalization necessarily limits the space of items that users are exposed to. Therefore, prior work notes a tension between personalization and users' ability for discovering novel items. While discovery of novel items in personalization setups may be resolved through search result diversification, these approaches do little to allow user control over personalization. Therefore, in this paper, we introduce an approach for controllable personalized search. Our model, CtrlCE presents a novel cross-encoder model augmented with an editable memory constructed from users historical items. Our proposed memory augmentation allows cross-encoder models to condition on large amounts of historical user data and supports interaction from users permitting control over personalization. Further, controllable personalization for search must account for queries which don't require personalization, and in turn user control. For this, we introduce a calibrated mixing model which determines when personalization is necessary. This allows system designers using CtrlCE to only obtain user input for control when necessary. In multiple datasets of personalized search, we show CtrlCE to result in effective personalization as well as fulfill various key goals for controllable personalized search.</description><subject>Computer Science - Computation and Language</subject><subject>Computer Science - Information Retrieval</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNpjYJA0NNAzsTA1NdBPLKrILNMzMjE01DMwMrc04GRw8U3NzS-qVHAsTc9NzStJTVFwLsovLtZNzUvOT0ktKlZIyy9ScM7PKynKz8lJTMpJVQgAiubnJeZkVgEVB6cmFiVn8DCwpiXmFKfyQmluBnk31xBnD12wffEFRZm5iUWV8SB748H2GhNWAQAcMziR</recordid><startdate>20241104</startdate><enddate>20241104</enddate><creator>Mysore, Sheshera</creator><creator>Dhanania, Garima</creator><creator>Patil, Kishor</creator><creator>Kallumadi, Surya</creator><creator>McCallum, Andrew</creator><creator>Zamani, Hamed</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20241104</creationdate><title>Memory Augmented Cross-encoders for Controllable Personalized Search</title><author>Mysore, Sheshera ; Dhanania, Garima ; Patil, Kishor ; Kallumadi, Surya ; McCallum, Andrew ; Zamani, Hamed</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-arxiv_primary_2411_027903</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Computation and Language</topic><topic>Computer Science - Information Retrieval</topic><toplevel>online_resources</toplevel><creatorcontrib>Mysore, Sheshera</creatorcontrib><creatorcontrib>Dhanania, Garima</creatorcontrib><creatorcontrib>Patil, Kishor</creatorcontrib><creatorcontrib>Kallumadi, Surya</creatorcontrib><creatorcontrib>McCallum, Andrew</creatorcontrib><creatorcontrib>Zamani, Hamed</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Mysore, Sheshera</au><au>Dhanania, Garima</au><au>Patil, Kishor</au><au>Kallumadi, Surya</au><au>McCallum, Andrew</au><au>Zamani, Hamed</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Memory Augmented Cross-encoders for Controllable Personalized Search</atitle><date>2024-11-04</date><risdate>2024</risdate><abstract>Personalized search represents a problem where retrieval models condition on historical user interaction data in order to improve retrieval results. However, personalization is commonly perceived as opaque and not amenable to control by users. Further, personalization necessarily limits the space of items that users are exposed to. Therefore, prior work notes a tension between personalization and users' ability for discovering novel items. While discovery of novel items in personalization setups may be resolved through search result diversification, these approaches do little to allow user control over personalization. Therefore, in this paper, we introduce an approach for controllable personalized search. Our model, CtrlCE presents a novel cross-encoder model augmented with an editable memory constructed from users historical items. Our proposed memory augmentation allows cross-encoder models to condition on large amounts of historical user data and supports interaction from users permitting control over personalization. Further, controllable personalization for search must account for queries which don't require personalization, and in turn user control. For this, we introduce a calibrated mixing model which determines when personalization is necessary. This allows system designers using CtrlCE to only obtain user input for control when necessary. In multiple datasets of personalized search, we show CtrlCE to result in effective personalization as well as fulfill various key goals for controllable personalized search.</abstract><doi>10.48550/arxiv.2411.02790</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2411.02790
ispartof
issn
language eng
recordid cdi_arxiv_primary_2411_02790
source arXiv.org
subjects Computer Science - Computation and Language
Computer Science - Information Retrieval
title Memory Augmented Cross-encoders for Controllable Personalized Search
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-22T20%3A00%3A43IST&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=Memory%20Augmented%20Cross-encoders%20for%20Controllable%20Personalized%20Search&rft.au=Mysore,%20Sheshera&rft.date=2024-11-04&rft_id=info:doi/10.48550/arxiv.2411.02790&rft_dat=%3Carxiv_GOX%3E2411_02790%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