Latent Unexpected Recommendations release_oxnronmcp5bm7h5sllm3axkryu

by Pan Li, Alexander Tuzhilin

Entity Metadata (schema)

abstracts[] {'sha1': '2381acf187470cd83b7c174e0b3365971c1a8ffe', 'content': 'Unexpected recommender system constitutes an important tool to tackle the\nproblem of filter bubbles and user boredom, which aims at providing unexpected\nand satisfying recommendations to target users at the same time. Previous\nunexpected recommendation methods only focus on the straightforward relations\nbetween current recommendations and user expectations by modeling\nunexpectedness in the feature space, thus resulting in the loss of accuracy\nmeasures in order to improve unexpectedness performance. Contrast to these\nprior models, we propose to model unexpectedness in the latent space of user\nand item embeddings, which allows to capture hidden and complex relations\nbetween new recommendations and historic purchases. In addition, we develop a\nnovel Latent Closure (LC) method to construct hybrid utility function and\nprovide unexpected recommendations based on the proposed model. Extensive\nexperiments on three real-world datasets illustrate superiority of our proposed\napproach over the state-of-the-art unexpected recommendation models, which\nleads to significant increase in unexpectedness measure without sacrificing any\naccuracy metric under all experimental settings in this paper.', 'mimetype': 'text/plain', 'lang': 'en'}
container
container_id
contribs[] {'index': 0, 'creator_id': None, 'creator': None, 'raw_name': 'Pan Li', 'given_name': None, 'surname': None, 'role': 'author', 'raw_affiliation': None, 'extra': None}
{'index': 1, 'creator_id': None, 'creator': None, 'raw_name': 'Alexander Tuzhilin', 'given_name': None, 'surname': None, 'role': 'author', 'raw_affiliation': None, 'extra': None}
ext_ids {'doi': None, 'wikidata_qid': None, 'isbn13': None, 'pmid': None, 'pmcid': None, 'core': None, 'arxiv': '2007.13280v1', 'jstor': None, 'ark': None, 'mag': None, 'doaj': None, 'dblp': None, 'oai': None, 'hdl': None}
files[] {'state': 'active', 'ident': '22g2f2mx6vg77lffpvhz7uueoe', 'revision': 'd3f9d6e1-4e6d-4720-b42b-74db9e56296c', 'redirect': None, 'extra': None, 'edit_extra': None, 'size': 7321178, 'md5': 'e5eb44f3657d535c373c4882377707e8', 'sha1': '455c4fad47ce2f02a9e83cab27d747605b04128e', 'sha256': '9521d39b9d6340fb84ea98892d6fefa42a005201a98ab926fd69cebb5d0ab004', 'urls': [{'url': 'https://arxiv.org/pdf/2007.13280v1.pdf', 'rel': 'repository'}, {'url': 'https://web.archive.org/web/20200929233028/https://arxiv.org/pdf/2007.13280v1.pdf', 'rel': 'webarchive'}], 'mimetype': 'application/pdf', 'content_scope': None, 'release_ids': ['oxnronmcp5bm7h5sllm3axkryu'], 'releases': None}
filesets []
issue
language en
license_slug ARXIV-1.0
number
original_title
pages
publisher
refs []
release_date 2020-07-27
release_stage submitted
release_type article
release_year 2020
subtitle
title Latent Unexpected Recommendations
version v1
volume
webcaptures []
withdrawn_date
withdrawn_status
withdrawn_year
work_id 3ifdtpo52zcdvfpvqp53o3zo5u
As JSON via API

Extra Metadata (raw JSON)

arxiv.base_id 2007.13280
arxiv.categories ['cs.IR', 'cs.LG']
arxiv.comments Accepted at ACM TIST