Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 40 of the 40 references it cites were found in a registry or an index; 4 of the 31 sources its claims cite were checked for support.

Revamp: Enhancing Accessible Information Seeking Experience of Online Shopping for Blind or Low Vision Users

Open access
Wang, Ruolin, Chen, Zixuan, Zhang, Mingrui Ray, Li, Zhaoheng, Liu, Zhixiu, Dang, Zihan, Yu, Chun, Chen, Xiang 'Anthony' · 2021 · Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems

Abstract

Online shopping has become a valuable modern convenience, but blind or low vision (BLV) users still face significant challenges using it, because of: 1) inadequate image descriptions and 2) the inability to filter large amounts of information using screen readers. To address those challenges, we propose Revamp, a system that leverages customer reviews for interactive information retrieval. Revamp is a browser integration that supports review-based question-answering interactions on a reconstructed product page. From our interview, we identified four main aspects (color, logo, shape, and size) that are vital for BLV users to understand the visual appearance of a product. Based on the findings, we formulated syntactic rules to extract review snippets, which were used to generate image descriptions and responses to users’ queries. Evaluations with eight BLV users showed that Revamp 1) provided useful descriptive information for understanding product appearance and 2) helped the participants locate key information efficiently.

Overview

CoChat Summary

TL;DR

Revamp is a browser extension for Amazon that reconstructs product pages and uses customer reviews to generate image descriptions and answer blind or low vision (BLV) users’ visual questions. Evaluations with BLV participants found it improved understanding of product appearance and information-seeking efficiency compared with current Amazon use.

Key findings

  • In a formative study with 20 BLV participants, the authors collected 168 questions; 57% were visual and 43% non-visual, and four main visual aspects were identified: color, logo, shape, and size.
  • Rule-selected review snippets gained 85% of 144 votes from eight BLV participants over snippets from Amazon’s existing “Have a Question” section.
  • On 45 Amazon top-seller products, the rules directly provided visual information for 54 questions/image descriptions in Clothing, Shoes & Jewelry, 40 in Home & Kitchen, and 29 in Electronics.
  • In system evaluation with eight BLV users, Revamp significantly outperformed current practice on all rated statements: understanding product appearance was 5.88 (1.25) vs. 3.00 (0.76), locating needed information was 5.50 (1.07) vs. 4.75 (0.89), and participants rated Revamp as a helpful independent-shopping alternative at 6.50 (0.53).

Methods

  • Three study groups: a two-stage formative study with 20 BLV participants (30-minute interviews; second stage with 10 participants), rule evaluation with eight BLV participants plus two sighted researchers rating 45 products, and system evaluation with eight BLV participants browsing six products for about 40 minutes.
  • Revamp was implemented as a Chrome extension on Amazon: it scrapes product, review, and Q&A data; simplifies the DOM; extracts query keywords; applies rule-based filtering and ranking; performs sentiment classification; and generates alt-text, positive/negative review lists, and a review summary.
  • Qualitative data were analyzed with reflexive thematic analysis and affinity diagramming; Likert-scale ratings were analyzed with the Wilcoxon test.

Limitations

  • The paper states that the system evaluation did not counterbalance conditions, all participants used Amazon first and Revamp second, there was possible bias from comparing a new system to an old one, and only eight participants and a limited number of products were involved.
  • The paper also states that the rules cover only four basic visual attributes (color, logo, shape, size), not high-level visual concepts, and that performance depends on review quality and availability.
  • Evident from the text: generated appearance descriptions may be too subjective because they contain only one snippet per visual aspect, and one participant preferred not to use any review filter in case important information was missed.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1145/3411764.3445547

Cites

40 of 40 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Image 1
From the PDF
Figure 2 System overview of how Revamp responds to users' query. It first extract the keywords and process the query by its kind. After filtering out irrelevant reviews and reranking, Revamp uses the shortest candidate among the top-3 reviews to generate the visual description of a product and relevant answers to the query.
Figure 2 System overview of how Revamp responds to users' query. It first extract the keywords and process the query by its kind. After filtering out irrelevant reviews and reranking, Revamp uses the shortest candidate among the top-3 reviews to generate the visual description of a product and relevant answers to the query.From the PDF
Figure 3 An example questionnaire for BLV users to vote for which one of the review snippets (retrieved by our rules or by default ranking of search result) best answers visual attributes questions.
Figure 3 An example questionnaire for BLV users to vote for which one of the review snippets (retrieved by our rules or by default ranking of search result) best answers visual attributes questions.From the PDF
Figure 4 Performance of image description and visual questions on four basic attributes including shape, logo, size, color. Two sighted research team members rated the quality by three levels on 15 Amazon top-seller products in each of the three main categories. Results showed that our rules can directly provide visual info for 54 questions or image descriptions in Clothing, Shoes & Jewelry; 40 in Home & Kitchen; 29 in Electronics. We also included 'Not Applicable' because of no relevant reviews or too few informative reviews.
Figure 4 Performance of image description and visual questions on four basic attributes including shape, logo, size, color. Two sighted research team members rated the quality by three levels on 15 Amazon top-seller products in each of the three main categories. Results showed that our rules can directly provide visual info for 54 questions or image descriptions in Clothing, Shoes & Jewelry; 40 in Home & Kitchen; 29 in Electronics. We also included 'Not Applicable' because of no relevant reviews or too few informative reviews.From the PDF
Image 5
From the PDF
Image 6
From the PDF
Image 7
From the PDF
Figure 6 Subjective ratings on statements comparing current practice with experience of using Revamp. Revamp demonstrates a clear advantage in the experience of understanding the product appearance.
Figure 6 Subjective ratings on statements comparing current practice with experience of using Revamp. Revamp demonstrates a clear advantage in the experience of understanding the product appearance.From the PDF

Verification

Claims the paper makes

4 Valid · 8 Not valid · 11 Not checked · 82 Own results and interpretations

46 more claims

Create a free account to see all 105 claims and to verify the 2 references still open.

See all claims — free account
  • It may state two separate claims as one. 24
  • Probably not a substantial claim. 35

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

2 of 2

Citations

Cites

Show all 39 references

Research works printed but not linked: 3

  1. Fangtao Li, Chao Han, Minlie Huang, Xiaoyan Zhu, Ying-Ju Xia, Shu Zhang, and Hao Yu. 2010. Structure-aware review mining and summarization. In Proceedings of the 23rd international conference on computational linguistics. Association for Computational Linguistics, Association for Computational Linguistics, Stroudsburg, PA, 653–661.
  2. Quang Nguyen. 2012. Detecting experience revealing sentences in product reviews. University of Amsterdam 1, 1 (2012), 1–78.
  3. 10.18653/v1 · doi:10.18653/v1

Cited by

See all 42

Metadata

Authors
Wang, R., Chen, Z., Zhang, M. R., Li, Z., Liu, Z., Dang, Z., Yu, C., & Chen, X. '.
Affiliations
  • University of California, Los Angeles (US)
  • University of Washington (US)
  • Tsinghua University (CN)
  • Stanford University (US)
Venue
Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems
Publisher
ACM
Published
2021-05-06
Pages
1-14
OpenAlex
W3128961190
Cited by
42
References
39
Citations by year
2026: 8 · 2025: 10 · 2024: 4 · 2023: 11 · 2022: 9
FWCI
7.88
Type
conference paper
Language
en
Open access
Locations
  • CHI Conference on Human Factors in Computing Systems (publishedVersion, cc-by)
  • arXiv (Cornell University) (submittedVersion)
  • arXiv (Cornell University)
Topics
  • Social Sciences › Human Factors and Ergonomics › Digital Accessibility for Disabilities
  • Health Sciences › Health Professions › Occupational Therapy › Assistive Technology in Communication and Mobility
  • Physical Sciences › Computer Science › Human-Computer Interaction › Usability and User Interface Design
Keywords
blind and visually impaired users, screen readers, customer reviews, image description generation, syntactic rules, browser extension, information retrieval

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Revamp: Enhancing Accessible Information Seeking Experience of Online Shopping for Blind or Low Vision Users. https://app.cochat.ai/papers/10.1145/3411764.3445547
The paper
Wang, R., Chen, Z., Zhang, M. R., Li, Z., Liu, Z., Dang, Z., Yu, C., & Chen, X. '. (2021). Revamp: Enhancing Accessible Information Seeking Experience of Online Shopping for Blind or Low Vision Users. Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems. https://doi.org/10.1145/3411764.3445547

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.