Pith. sign in

REVIEW 3 major objections 7 minor 96 references

Adapting Online Customer Reviews for Blind Users: A Case Study of Restaurant Reviews

T0 review · 3 major / 7 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A screen-reader browser extension re-organizes restaurant reviews by aspect and sentiment, letting blind users find the good and bad without listening through repetitive reviews.

desk verdict A well-motivated accessibility tool with a strong interview study and a plausible usability win, but the statistics and LLM evaluation need tightening before the decision-making claim can be trusted. read the letter →

arxiv 2506.04865 v1 pith:LZ6SV4SU submitted 2025-06-05 cs.HC

Mohan Sunkara , Akshay Kolgar Nayak , Sandeep Kalari , Yash Prakash , Sampath Jayarathna , Hae-Na Lee , Vikas Ashok This is my paper
classification cs.HC
keywords blindusersscreenreaderaccessibilityrestaurantreviewslargelanguagemodelaspect-basedsentimentanalysisfocusedsummarizationGoogleMaps
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Online restaurant reviews are hard for blind users, the paper argues, not because the content is unavailable but because screen readers narrate it in one long linear stream and the reviews repeat themselves. The authors test this by building QuickCue, a Chrome extension that uses GPT-4 to tag each review with aspect–sentiment pairs (food, ambiance, customer service, pricing, and hygiene, each positive or negative) and then writes a short focused summary for every pair. In a within-subject study with 10 blind participants, QuickCue raised the average System Usability Scale score from 63.25 with the default Google Maps screen reader experience to 81.5 ($F=45.03$, $p=2.72\times10^{-6}$), and cut NASA-TLX workload from 62.09 to 38.37 ($F=99.27$, $p=9.45\times10^{-9}$). The authors conclude that reorganizing reviews before narration, rather than making the screen reader faster, is a workable route to accessible review browsing.

What carries the argument

The load-bearing mechanism is a two-stage LLM pipeline. Stage one, joint aspect–sentiment classification, uses a modified Clue and Reasoning (CARP) prompt so GPT-4 outputs every (aspect, sentiment) pair found in a review instead of a single label; this handles mixed-opinion reviews like 'food was good, service slow.' Stage two, focused summarization, uses Directional Stimulus Prompting (DSP) with the target aspect and sentiment supplied as directional stimuli, producing bullet-point summaries that exclude unrelated topics. The accessible interface is an ARIA-annotated accordion injected into Google Maps, navigable with TAB, ENTER, and ESCAPE, which keeps the interaction within the user's existing screen reader and hotkey habits.

What would settle it

Take QuickCue's classifier and summarizer to a fresh set of, say, 100 recently posted restaurant reviews from regions and cuisines outside the 50 used for evaluation, and measure average F1 and factuality. A clear drop below roughly 0.81 F1 or 7.9/10 factuality, or a failure to reproduce the SUS and NASA-TLX advantages in a preregistered replication with more than 10 participants, would show the result is tied to the small test set and the specific sample rather than to the reorganized presentation itself.

Watch

Extended reading notes

Core claim

QuickCue replaces the flat list of restaurant reviews with a three-level structure: five aspect buttons, positive and negative summaries under each aspect, and the original supporting reviews under each summary. Both the grouping and the summarization are performed by GPT-4, using a modified Clue and Reasoning prompt for joint classification (one review can mention several aspects with mixed sentiments) and a Directional Stimulus prompt that forces each summary to stay on one aspect and one sentiment. In a counterbalanced within-subject study with 10 JAWS screen reader users comparing two restaurants, the QuickCue condition outperformed the status quo on usability (SUS 81.5 versus 63.25, $F=45.03$, $p=2.72\times10^{-6}$) and on perceived workload (NASA-TLX 38.37 versus 62.09, $F=99.27$, $p=9.45\times10^{-9}$). The paper presents this as evidence that thematic, sentiment-segregated presentation is what makes reviews perusable for blind users, with several participants relying on the summaries alone and skipping the original reviews.

Load-bearing premise

QuickCue's benefit depends on GPT-4 reliably classifying reviews into aspect–sentiment pairs and writing factual focused summaries; the paper's own checks show average F1 of 0.81 and factuality of 7.9 out of 10 on 50-item test sets, so if accuracy degrades on unseen restaurants or review styles, the usability gains found with 10 participants may not transfer.

Editorial extensions

If this is right

  • If the user-study results transfer, the presentation layer, not the review text itself, is the main obstacle: reorganizing the same reviews by aspect and sentiment raised perceived usability by more than 18 SUS points.
  • Blind diners can make more informed choices without extra effort; seven participants said QuickCue would let them try new dishes instead of ordering the same familiar items.
  • The two-stage pipeline is claimed to generalize beyond restaurants: changing the aspect vocabulary and prompt templates should adapt it to e-commerce and other review-based platforms, because the architecture is modular.
  • Because several participants skipped the original reviews entirely and listened only to summaries, the design direction is reducing listening volume rather than increasing narration speed.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A consequence the paper leaves untested is whether smaller or cheaper models than GPT-4 can carry the same gains; the classifier and summarizer scores are strong enough that a cost-performance study on alternative LLMs would be a natural next step.
  • The evaluation used only 50 reviews for classification and 50 examples for summaries, so a larger held-out sample across cuisines and languages is the direct way to check whether the 0.81 F1 and 7.9/10 factuality hold where QuickCue would actually run.
  • The paper's own qualitative finding, that users repeat the same aspect search when moving from one restaurant to the next, suggests personalized persistent aspect preferences would be a high-value addition, and the authors mention it only as future work.
  • Because all 10 user-study participants used JAWS on Windows, the result says nothing yet about NVDA or VoiceOver users; a replication with other screen readers would test the paper's claim that the interface is screen-reader-agnostic.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 7 minor

Summary. The paper presents QuickCue, a browser extension that augments Google Maps restaurant reviews for blind screen reader users by organizing reviews into aspect-sentiment groups and generating focused summaries with GPT-4. The authors first report a semi-structured interview study with 30 blind participants identifying pain points such as listening fatigue, redundancy, and difficulty locating specific information. They then design QuickCue using a modified CARP prompt for joint aspect-sentiment classification and a DSP prompt for focused summarization, and evaluate these components on small self-constructed datasets. Finally, they report a within-subject user study with 10 blind participants in which QuickCue yielded significantly higher SUS scores (63.25 vs. 81.5) and significantly lower NASA-TLX scores (62.09 vs. 38.37) than the default Google Maps screen reader experience, along with qualitative feedback supporting perceived usability benefits.

Significance. If the reported usability and workload benefits are borne out, this is a practically valuable contribution to accessibility research: it addresses a real, underserved interaction problem, and the user study is an independent comparison against the status quo rather than a circular evaluation of the system's own components. The interview findings and the modular design also provide a useful foundation for other review platforms. However, the strength of the central claim is tempered by an invalid statistical analysis of the within-subject data, and the paper's decision-making claims outrun the evidence: the LLM component evaluations are small and lack reliability measures, and the user study did not verify the factual quality of the summaries actually presented to participants. These issues are fixable through re-analysis and careful claim-scoping, so the contribution remains defensible.

major comments (3)
  1. [§5.4.1, §5.4.2] The SUS and TLX comparisons are analyzed with one-way ANOVA on a within-subject design. Since each participant contributes paired observations in both conditions, the independence assumption of ANOVA is violated, and the reported F and p values (F=45.03, p=2.72e-6; F=99.27, p=9.45e-9) are not valid as reported. Please re-analyze with a paired-samples t-test or repeated-measures ANOVA and report effect sizes (e.g., Cohen's dz) and confidence intervals. Given the magnitude of the observed differences, the substantive conclusion may survive, but the statistical support must be recomputed before the headline claim can be accepted.
  2. [§4.2, §4.3, §6.1] The claim that QuickCue supports "more informed decisions" rests on the factual quality of the LLM-generated aspect-sentiment classifications and focused summaries. The component evaluations use 50 manually annotated reviews for joint classification and 50 self-constructed examples for summarization, with no inter-annotator reliability measure and no human summarization baseline; moreover, the user study did not verify whether the summaries shown to participants were factually correct. As the authors acknowledge in §6.1, the evaluation was mostly qualitative and lacking quantitative metrics such as task completion time or error rates. The study therefore demonstrates a perceived usability and workload benefit, but not an objective decision-making benefit. Please either remove or explicitly scope the decision-making claims in the abstract and conclusion to self-reported perceptions, or add a behavioral measure of decision accuracy or comprehension.
  3. [§5.4.3, §7] Qualitative statements such as "QuickCue would enable them to make more informed decisions" are treated in the abstract and conclusion as evidence of a decision-making benefit. Since these are exit-interview self-reports rather than measures of decision quality, they should be reported as perceived benefits, not as demonstrated improvements in decision outcomes. The conclusion's phrase "thereby enhancing decision-making" should be revised to reflect the evidence actually collected.
minor comments (7)
  1. [Abstract] The abstract refers to "QuickQue" instead of "QuickCue".
  2. [§4.2] The text reads "large large model" and contains missing spaces between words in the prompt-engineering description; please correct these typos.
  3. [Figure 3] The plots show only means; please add error bars or individual participant points so the spread of SUS and TLX scores is visible.
  4. [Table 1] The "Age of Vision Loss" column contains "NA" for P7 and P8; please clarify whether this means congenital, unknown, or non-applicable.
  5. [§4.2] Please state the specific GPT-4 model version and access date, since prompt-based evaluation results are version-sensitive.
  6. [§4.2] The frequency breakdown sums to 121 labels across 50 reviews; please state explicitly that reviews can carry multiple aspect-sentiment labels to avoid confusion.
  7. [§5.2] Please report the JAWS screen reader version used in the study for reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: QuickCue's central usability claim rests on an independent within-subject comparison against the default Google Maps screen-reader experience, and the LLM component evaluations are separate from the user-study outcome.

full rationale

The paper's central claim is an empirical usability comparison: in a within-subject study, 10 blind participants rated QuickCue versus the status quo screen reader on Google Maps, with SUS rising from 63.25 to 81.5 and NASA-TLX falling from 62.09 to 38.37 (Section 5.4). This outcome is measured, not derived from QuickCue's construction; the baseline is an external system (the default Google Maps interface), so the comparison is self-contained. The LLM components are evaluated separately against manually annotated test sets (50 reviews for joint classification, Section 4.2; 50 examples with handcrafted summaries for focused summarization, Section 4.3). Those evaluations are in-house, but they are not predictions and do not feed the user-study numbers; they are standard component-level checks. The paper's self-citations appear only in related-work and prior-tool contexts (e.g., [9, 34, 51, 67, 68, 77]) and are not load-bearing premises: the motivating interview study is original data reported in Section 3, and the prompting methods CARP [76] and DSP [55] are external. The stated limitations (Section 6.1: 'our evaluation was mostly qualitative, lacking quantitative metrics such as task completion time or error rates'; small 50-example component datasets) weaken the support for the broader 'informed decisions' claim, but they are validity concerns, not circularity. No equation-level reduction, fitted-parameter-as-prediction, or self-citation chain forces the reported results. Therefore no significant circularity is present.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new theoretical entities. Its assumptions are the reliability of the LLM, the validity of a simple ANOVA on repeated measures, the representativeness of a small convenience sample, and the accuracy of self-constructed annotations. The free parameters are design choices (aspect set and few-shot example counts) rather than fitted numerical constants.

free parameters (3)
  • Number of few-shot examples for CARP joint classification = 20
    Handcrafted by the authors based on prior work's recommendation of at least 16; not derived from a fitting procedure, but chosen by hand and affecting classifier performance.
  • Number of few-shot examples for DSP focused summarization = 10
    One handcrafted example per aspect-sentiment pair; chosen by the authors, not fitted.
  • Predefined aspect set = Food, Ambiance, Hygiene, Customer Service, Pricing
    Selected based on the interview study; this categorical choice determines the grouping structure and is not derived from data.
assumptions (4)
  • domain assumption GPT-4 produces accurate joint aspect-sentiment classifications and faithful focused summaries for restaurant reviews.
    The entire tool depends on LLM quality; the paper evaluates this on 50 self-annotated examples, not on independent benchmarks.
  • standard math One-way ANOVA can be applied to within-subject paired data.
    Used in Section 5.4.1 and 5.4.2; the independence assumption is violated for the repeated-measures design, which may inflate significance.
  • domain assumption The 10 JAWS users in the user study are representative of the broader blind screen reader population.
    Recruited by email lists and snowball sampling; all JAWS and Chrome proficient, limiting generalizability.
  • domain assumption Manual annotations of the ground truth and human ratings of summaries are accurate and unbiased.
    Ground truth for 50 reviews was annotated by one research assistant and verified by two others; summary ratings from 10 annotators may be influenced by the authors' own prompt design.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Adapting Online Customer Reviews for Blind Users: A Case Study of Restaurant Reviews." pith.science (2026). https://pith.science/paper/LZ6SV4SU

@misc{pith2026250604865,
  author       = {Pith},
  title        = {Pith review of: Adapting Online Customer Reviews for Blind Users: A Case Study of Restaurant Reviews},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LZ6SV4SU}},
  note         = {Machine review of arXiv:2506.04865}
}
read the original abstract

Online reviews have become an integral aspect of consumer decision-making on e-commerce websites, especially in the restaurant industry. Unlike sighted users who can visually skim through the reviews, perusing reviews remains challenging for blind users, who rely on screen reader assistive technology that supports predominantly one-dimensional narration of content via keyboard shortcuts. In an interview study, we uncovered numerous pain points of blind screen reader users with online restaurant reviews, notably, the listening fatigue and frustration after going through only the first few reviews. To address these issues, we developed QuickQue assistive tool that performs aspect-focused sentiment-driven summarization to reorganize the information in the reviews into an alternative, thematically-organized presentation that is conveniently perusable with a screen reader. At its core, QuickQue utilizes a large language model to perform aspect-based joint classification for grouping reviews, followed by focused summarizations within the groups to generate concise representations of reviewers' opinions, which are then presented to the screen reader users via an accessible interface. Evaluation of QuickQue in a user study with 10 participants showed significant improvements in overall usability and task workload compared to the status quo screen reader.

Figures

Figures reproduced from arXiv: 2506.04865 by the authors.

Figure 1
Figure 1. QuickCue’s interface hierarchy: (A) displays the default list of five aspects, (B) shows the positives and negatives of [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. A workflow schematic depicting QuickCue’s architecture. [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. (Left) SUS scores, (right) NASA-TLX scores. [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

96 extracted references · 69 canonical work pages

  1. [1]

    Julio Abascal, Myriam Arrue, and Xabier Valencia. 2019. Tools for web accessi- bility evaluation.Web accessibility: a foundation for research(2019), 479–503

    2019

  2. [2]

    Shadi Abou-Zahra. 2008. Web accessibility evaluation.Web accessibility: A foundation for research(2008), 79–106

    2008

  3. [3]

    Hayfa Y Abuaddous, Mohd Zalisham Jali, and Nurlida Basir. 2016. Web ac- cessibility challenges.International Journal of Advanced Computer Science and Applications7, 10 (2016)

    2016

  4. [4]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Floren- cia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. 2023. Gpt-4 technical report.arXiv preprint arXiv:2303.08774 (2023)

    arXiv 2023

  5. [5]

    Toufique Ahmed and Premkumar Devanbu. 2022. Few-shot training llms for project-specific code-summarization. InProceedings of the 37th IEEE/ACM inter- national conference on automated software engineering. 1–5

    2022

  6. [6]

    Explosion AI. 2024. spaCy: Industrial-Strength Natural Language Processing. https://spacy.io/ Accessed: March 3, 2025

    2024

  7. [7]

    Apple Inc. 2022. Accessibility - Vision - Apple. https://www.apple.com/ accessibility/vision/. Adapting Online Customer Reviews for Blind Users: A Case Study of Restaurant Reviews W4A ’25, April 28-29, 2025, Sydney, Australia

    2022

  8. [8]

    Jinat Ara, Md Toufique Hasan, Abdullah Al Omar, and Hanif Bhuiyan. 2020. Understanding customer sentiment: Lexical analysis of restaurant reviews. In 2020 IEEE Region 10 Symposium (TENSYMP). IEEE, 295–299

    2020

Show all 96 references
  1. [9]

    Vikas Ashok, Syed Masum Billah, Yevgen Borodin, and IV Ramakrishnan. 2019. Auto-suggesting browsing actions for personalized web screen reading. InPro- ceedings of the 27th ACM Conference on User Modeling, Adaptation and Personal- ization. 252–260

    2019

  2. [10]

    Vikas Ashok, Yevgen Borodin, Svetlana Stoyanchev, Yuri Puzis, and IV Ramakrish- nan. 2014. Wizard-of-Oz evaluation of speech-driven web browsing interface for people with vision impairments. InProceedings of the 11th Web for All Conference. 1–9

    2014

  3. [11]

    Vikas Ashok, Yury Puzis, Yevgen Borodin, and IV Ramakrishnan. 2017. Web screen reading automation assistance using semantic abstraction. InProceedings of the 22nd International Conference on Intelligent User Interfaces. 407–418

    2017

  4. [12]

    Ali Selman Aydin, Shirin Feiz, Vikas Ashok, and IV Ramakrishnan. 2020. Sail: Saliency-driven injection of aria landmarks. InProceedings of the 25th international conference on intelligent user interfaces. 111–115

    2020

  5. [13]

    Jeffrey P Bigham, Anna C Cavender, Jeremy T Brudvik, Jacob O Wobbrock, and Richard E Ladner. 2007. WebinSitu: a comparative analysis of blind and sighted browsing behavior. InProceedings of the 9th International ACM SIGACCESS Conference on Computers and Accessibility. 51–58

    2007

  6. [14]

    Nilton Bila, Troy Ronda, Iqbal Mohomed, Khai N Truong, and Eyal De Lara. 2007. Pagetailor: reusable end-user customization for the mobile web. InProceedings of the 5th international conference on Mobile systems, applications and services. 16–29

    2007

  7. [15]

    Syed Masum Billah, Vikas Ashok, Donald E Porter, and IV Ramakrishnan. 2017. Speed-dial: A surrogate mouse for non-visual web browsing. InProceedings of the 19th International ACM SIGACCESS Conference on Computers and Accessibility. 110–119

    2017

  8. [16]

    Steven Bird, Ewan Klein, and Edward Loper. 2024. Natural Language Toolkit (NLTK). https://www.nltk.org/ Accessed: March 3, 2025

    2024

  9. [17]

    James V. Bradley. 1958. Complete Counterbalancing of Immediate Se- quential Effects in a Latin Square Design.J. Amer. Statist. Assoc. 53, 282 (1958), 525–528. https://doi.org/10.1080/01621459.1958.10501456 arXiv:https://amstat.tandfonline.com/doi/pdf/10.1080/01621459.1958.10501456

    1958

  10. [18]

    Giorgio Brajnik. 2004. Comparing accessibility evaluation tools: a method for tool effectiveness.Universal access in the information society3 (2004), 252–263

    2004

  11. [19]

    Anke M Brock, Philippe Truillet, Bernard Oriola, Delphine Picard, and Christophe Jouffrais. 2015. Interactivity improves usability of geographic maps for visually impaired people.Human–Computer Interaction30, 2 (2015), 156–194

    2015

  12. [20]

    John Brooke. 1996. Sus: a “quick and dirty’usability.Usability evaluation in industry189, 3 (1996)

    1996

  13. [21]

    Ben Caldwell, Michael Cooper, Loretta Guarino Reid, Gregg Vanderheiden, Wendy Chisholm, John Slatin, and Jason White. 2008. Web content accessibility guide- lines (WCAG) 2.0.WWW Consortium (W3C)290 (2008), 1–34

    2008

  14. [22]

    Harrison Chase. 2022. LangChain: Building Applications with LLMs through Composability. GitHub Repository. https://github.com/langchain-ai/langchain

    2022

  15. [23]

    Garima Chhikara, Anurag Sharma, V Gurucharan, Kripabandhu Ghosh, and Abhijnan Chakraborty. 2024. LaMSUM: Amplifying Voices Against Harassment through LLM Guided Extractive Summarization of User Incident Reports.arXiv preprint arXiv:2406.15809(2024)

    2024 arXiv

  16. [24]

    Wendy Chisholm, Gregg Vanderheiden, and Ian Jacobs. 2001. Web content accessibility guidelines 1.0.Interactions8, 4 (2001), 35–54

    2001

  17. [25]

    World Wide Web Consortium et al. 1999. Web content accessibility guidelines 1.0. (1999)

    1999

  18. [26]

    Anupam Dash, Dongsong Zhang, and Lina Zhou. 2021. Personalized ranking of online reviews based on consumer preferences in product features.International Journal of Electronic Commerce25, 1 (2021), 29–50

    2021

  19. [27]

    Kevin De Boeck, Jenno Verdonck, Michiel Willocx, Jorn Lapon, and Vincent Naessens. 2022. Reviewing review platforms: a privacy perspective. InProceedings of the 17th International Conference on A vailability, Reliability and Security. 1–10

    2022

  20. [28]

    Flask-RESTful Developers. 2024. Flask-RESTful Documentation. https://flask- restful.readthedocs.io/en/latest/ Accessed: March 3, 2025

    2024

  21. [29]

    Google Developers. 2024. Chrome Extensions Documentation. https://developer. chrome.com/docs/extensions/ Accessed: March 3, 2025

    2024

  22. [30]

    Tiago do Carmo Nogueira, Deller James Ferreira, Sérgio Teixeira de Carvalho, Luciana de Oliveira Berretta, and Mycke R Guntijo. 2019. Comparing sighted and blind users task performance in responsive and non-responsive web design. Knowledge and Information Systems58 (2019), 319–339

    2019

  23. [31]

    Wafaa S El-Kassas, Cherif R Salama, Ahmed A Rafea, and Hoda K Mohamed

  24. [32]

    Mohammed Fakrudeen, Sufian Yousef, and Mahdi H Miraz. 2017. Finger Based Technique (FBT): An Innovative System for Improved Usability for the Blind Users’ Dynamic Interaction with Mobile Touch Screen Devices.arXiv preprint arXiv:1708.05073(2017)

    2017 arXiv

  25. [33]

    Limin Fang. 2022. The Effects of Online Review Platforms on Restaurant Revenue, Consumer Learning, and Welfare.Manag. Sci.68 (2022), 8116–8143. https: //api.semanticscholar.org/CorpusID:246595766

    2022

  26. [34]

    Javedul Ferdous, Hae-Na Lee, Sampath Jayarathna, and Vikas Ashok. 2022. InSup- port: Proxy Interface for Enabling Efficient Non-Visual Interaction with Web Data Records. In27th International Conference on Intelligent User Interfaces. 49–62

    2022

  27. [35]

    Javedul Ferdous, Sami Uddin, and Vikas Ashok. 2021. Semantic table-of-contents for efficient web screen reading. InProceedings of the 36th Annual ACM Symposium on Applied Computing. 1941–1949

    2021

  28. [36]

    Raffaele Filieri. 2015. What makes online reviews helpful? A diagnosticity- adoption framework to explain informational and normative influences in e- WOM.Journal of business research68, 6 (2015), 1261–1270

    2015

  29. [37]

    Python Software Foundation. 2024. re — Regular expression operations. https: //docs.python.org/3/library/re.html Accessed: March 3, 2025

    2024

  30. [38]

    Freedom Scientific. 2022. JAWS ® – Freedom Scientific. https://www. freedomscientific.com/products/software/jaws/

    2022

  31. [39]

    Ioana Frîncu. 2023. In search of the perfect prompt. (2023)

    2023

  32. [40]

    Guilherme Corredato Guerino and Natasha Malveira Costa Valentim. 2020. Us- ability and user experience evaluation of natural user interfaces: a systematic mapping study.Iet Software14, 5 (2020), 451–467

    2020

  33. [41]

    Dorit Hadar-Shoval, Zohar Elyoseph, and Maya Lvovsky. 2023. The plasticity of ChatGPT’s mentalizing abilities: personalization for personality structures. Frontiers in Psychiatry14 (2023), 1234397

    2023

  34. [42]

    Hart and Lowell E

    Sandra G. Hart and Lowell E. Staveland. 1988. Development of NASA-TLX (Task Load Index): Results of Empirical and Theoretical Research. InHuman Mental Workload, Peter A. Hancock and Najmedin Meshkati (Eds.). Advances in Psychology, Vol. 52. North-Holland, 139 – 183. https://do...

    1988 doi

  35. [43]

    Xin Jin, Jonathan Larson, Weiwei Yang, and Zhiqiang Lin. 2023. Binary code summarization: Benchmarking chatgpt/gpt-4 and other large language models. arXiv preprint arXiv:2312.09601(2023)

    2023 arXiv

  36. [44]

    Elnaz Karamad. 2023. THE IMPACT OF ONLINE REVIEWS ON CONSUMER DECISION-MAKING: A SURVEY IN THE UKRAINIAN MARKET.Market econ- omy: modern management theory and practice(2023). https://api.semanticscholar. org/CorpusID:268133863

    2023

  37. [45]

    Madiha Khalid, Imran Ashraf, Arif Mehmood, Saleem Ullah, Maqsood Ahmad, and Gyu Sang Choi. 2020. GBSVM: sentiment classification from unstructured reviews using ensemble classifier.Applied Sciences10, 8 (2020), 2788

    2020

  38. [46]

    Aurangzeb Khan, Umair Younis, Alam Sher Kundi, Muhammad Zubair Asghar, Irfan Ullah, Nida Aslam, and Imran Ahmed. 2020. Sentiment classification of user reviews using supervised learning techniques with comparative opinion mining perspective. InAdvances in Computer Vision: Proc...

    2020

  39. [47]

    Prerna Khanna, IV Ramakrishnan, Shubham Jain, Xiaojun Bi, and Aruna Bala- subramanian. 2024. Hand Gesture Recognition for Blind Users by Tracking 3D Gesture Trajectory. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems. 1–15

    2024

  40. [48]

    Pooja Kherwa and Poonam Bansal. 2017. Latent semantic analysis: an approach to understand semantic of text. In2017 International Conference on Current Trends in Computer, Electrical, Electronics and Communication (CTCEEC). IEEE, 870–874

    2017

  41. [49]

    Shashank Kumar, Jeevitha Shree DV, and Pradipta Biswas. 2021. Comparing ten WCAG tools for accessibility evaluation of websites.Technology and Disability 33, 3 (2021), 163–185

    2021

  42. [50]

    Jonathan Lazar, Aaron Allen, Jason Kleinman, and Chris Malarkey. 2007. What frustrates screen reader users on the web: A study of 100 blind users.International Journal of human-computer interaction22, 3 (2007), 247–269

    2007

  43. [51]

    Hae-Na Lee and Vikas Ashok. 2021. Towards Enhancing Blind Users’ Interaction Experience with Online Videos via Motion Gestures. InProceedings of the 32nd ACM Conference on Hypertext and Social Media. 231–236

    2021

  44. [52]

    Hae-Na Lee and Vikas Ashok. 2022. Customizable Tabular Access to Web Data Records for Convenient Low-vision Screen Magnifier Interaction.ACM Transac- tions on Accessible Computing (TACCESS)15, 2 (2022), 1–22

    2022

  45. [53]

    Hae-Na Lee, Vikas Ashok, and IV Ramakrishnan. 2020. Rotate-and-Press: A Non- visual Alternative to Point-and-Click?. InHCI International 2020–Late Breaking Papers: Universal Access and Inclusive Design: 22nd HCI International Conference, HCII 2020, Copenhagen, Denmark, July 19...

    2020

  46. [54]

    Nao Li, Xiaoyu Yang, IpKin Anthony Wong, Rob Law, Jing Yang Xu, and Binru Zhang. 2023. Automating tourism online reviews: a neural network based aspect- oriented sentiment classification.Journal of Hospitality and Tourism Technology 14, 1 (2023), 1–20

    2023

  47. [55]

    Zekun Li, Baolin Peng, Pengcheng He, Michel Galley, Jianfeng Gao, and Xifeng Yan. 2024. Guiding large language models via directional stimulus prompting. Advances in Neural Information Processing Systems36 (2024)

    2024

  48. [56]

    2022.Sentiment analysis and opinion mining

    Bing Liu. 2022.Sentiment analysis and opinion mining. Springer Nature

    2022

  49. [57]

    Christina Low, Emma McCamey, Cole Gleason, Patrick Carrington, Jeffrey P Bigham, and Amy Pavel. 2019. Twitter a11y: A browser extension to describe images. InProceedings of the 21st International ACM SIGACCESS Conference on Computers and Accessibility. 551–553. W4A ’25, April ...

    2019

  50. [58]

    Michael Luca. 2016. Reviews, reputation, and revenue: The case of Yelp. com. Com (March 15, 2016). Harvard Business School NOM Unit Working Paper12-016 (2016)

    2016

  51. [59]

    Michael Luca. 2016. Reviews, Reputation, and Revenue: The Case of Yelp.Com. https://api.semanticscholar.org/CorpusID:14511907

    2016

  52. [60]

    Calvin Luy, Jeremy Law, Lily Ho, Richard Matheson, Tracey Cai, Anuradha Madugalla, and John Grundy. 2021. A toolkit for building more adaptable user interfaces for vision-impaired users. In2021 IEEE symposium on visual languages and human-centric computing (VL/HCC). IEEE, 1–5

    2021

  53. [61]

    Valentyn Melnyk, Vikas Ashok, Valentyn Melnyk, Yury Puzis, Yevgen Borodin, Andrii Soviak, and IV Ramakrishnan. 2015. Look ma, no aria: generic accessible interfaces for web widgets. InProceedings of the 12th International Web for All Conference. 1–4

    2015

  54. [62]

    Michael Nebeling, Maximilian Speicher, and Moira C Norrie. 2013. CrowdAdapt: enabling crowdsourced web page adaptation for individual viewing conditions and preferences. InProceedings of the 5th ACM SIGCHI symposium on Engineering interactive computing systems. 23–32

    2013

  55. [63]

    NV Access. 2022. NV Access. https://www.nvaccess.org/

    2022

  56. [64]

    Uran Oh, Hwayeon Joh, and YunJung Lee. 2021. Image accessibility for screen reader users: A systematic review and a road map.Electronics10, 8 (2021), 953

    2021

  57. [65]

    2012.Grounded theory

    Julianne S Oktay. 2012.Grounded theory. Oxford University Press

    2012

  58. [66]

    OpenAI. 2024. GPT-4 Model Documentation. https://platform.openai.com/docs/ models/gpt-4 Accessed: March 3, 2025

    2024

  59. [67]

    Yash Prakash, Akshay Kolgar Nayak, Mohan Sunkara, Sampath Jayarathna, Hae- Na Lee, and Vikas Ashok. 2024. All in One Place: Ensuring Usable Access to Online Shopping Items for Blind Users.Proceedings of the ACM on Human-Computer Interaction8, EICS (2024), 1–25

    2024

  60. [68]

    Yash Prakash, Mohan Sunkara, Hae-Na Lee, Sampath Jayarathna, and Vikas Ashok. 2023. AutoDesc: facilitating convenient perusal of web data items for blind users. InProceedings of the 28th International Conference on Intelligent User Interfaces. 32–45

    2023

  61. [69]

    Pradeep Racherla and Wesley Friske. 2012. Perceived ’usefulness’ of online consumer reviews: An exploratory investigation across three services categories. Electron. Commer. Res. Appl.11 (2012), 548–559. https://api.semanticscholar.org/ CorpusID:28895996

    2012

  62. [70]

    Johnny Saldaña. 2021. The coding manual for qualitative researchers. (2021)

    2021

  63. [71]

    Ather Sharif, Sanjana Shivani Chintalapati, Jacob O Wobbrock, and Katharina Reinecke. 2021. Understanding screen-reader users’ experiences with online data visualizations. InProceedings of the 23rd International ACM SIGACCESS Conference on Computers and Accessibility. 1–16

    2021

  64. [72]

    Daniel Siahaan, Achmad Maududie, Agus Subekti, Hotniar Siringoringo, et al

  65. [73]

    Beverley A Sparks and Victoria Browning. 2011. The impact of online reviews on hotel booking intentions and perception of trust.Tourism management32, 6 (2011), 1310–1323

    2011

  66. [74]

    Manjunath Srinivas, S Vamsi Krishna Reddy, Manoj NM, and Hatsuho Miyazawa

  67. [75]

    Amanda Stent, Matthew Marge, and Mohit Singhai. 2005. Evaluating evaluation methods for generation in the presence of variation. InInternational conference on intelligent text processing and computational linguistics. Springer, 341–351

    2005

  68. [76]

    Xiaofei Sun, Xiaoya Li, Jiwei Li, Fei Wu, Shangwei Guo, Tianwei Zhang, and Guoyin Wang. 2023. Text Classification via Large Language Models. InFindings of the Association for Computational Linguistics: EMNLP 2023, Houda Bouamor, Juan Pino, and Kalika Bali (Eds.). Association f...

    2023 doi

  69. [77]

    Mohan Sunkara, Yash Prakash, Hae-Na Lee, Sampath Jayarathna, and Vikas Ashok. 2023. Enabling customization of discussion forums for blind users.Pro- ceedings of the ACM on Human-Computer Interaction7, EICS (2023), 1–20

    2023

  70. [78]

    Oguzhan Tas and Farzad Kiyani. 2007. A survey automatic text summarization. PressAcademia Procedia5, 1 (2007), 205–213

    2007

  71. [79]

    Gemini Team, Rohan Anil, Sebastian Borgeaud, Jean-Baptiste Alayrac, Jiahui Yu, Radu Soricut, Johan Schalkwyk, Andrew M Dai, Anja Hauth, Katie Millican, et al. 2023. Gemini: a family of highly capable multimodal models.arXiv preprint arXiv:2312.11805(2023)

    2023 arXiv

  72. [80]

    Mike Thelwall. 2024. ChatGPT for complex text evaluation tasks.Journal of the Association for Information Science and Technology(2024)

    2024

  73. [81]

    Hugo Touvron, Thibaut Lavril, Gautier Izacard, Xavier Martinet, Marie-Anne Lachaux, Timothée Lacroix, Baptiste Rozière, Naman Goyal, Eric Hambro, Faisal Azhar, et al. 2023. Llama: Open and efficient foundation language models.arXiv preprint arXiv:2302.13971(2023)

    2023 arXiv

  74. [82]

    Ana Valdivia, M Victoria Luzón, and Francisco Herrera. 2017. Sentiment analysis on tripadvisor: Are there inconsistencies in user reviews?. InHybrid Artificial Intelligent Systems: 12th International Conference, HAIS 2017, La Rioja, Spain, June 21-23, 2017, Proceedings 12. Spr...

    2017

  75. [83]

    Damir Vandic, Lennart J Nederstigt, Flavius Frasincar, Uzay Kaymak, and Enzo Ido. 2024. A framework for approximate product search using faceted navigation and user preference ranking.Data & Knowledge Engineering149 (2024), 102241

    2024

  76. [84]

    Byron C Wallace, Sayantan Saha, Frank Soboczenski, and Iain J Marshall. 2021. Generating (factual?) narrative summaries of rcts: Experiments with neural multi- document summarization.AMIA Summits on Translational Science Proceedings 2021 (2021), 605

    2021

  77. [85]

    WebAIM. 2019. WebAIM: Screen Reader User Survey #8 Results. https://webaim. org/projects/screenreadersurvey8/

    2019

  78. [86]

    Jared Smith WebAIM. 2015. Screen reader user survey# 5 results.http://webaim. org/projects/screenreadersurvey6/.(2015)

    2015

  79. [87]

    World Wide Web Consortium (W3C). 2021. Accessible Rich Internet Applications (WAI-ARIA) 1.2. https://www.w3.org/TR/wai-aria-1.2/. Accessed: 2025-02-14

    2021

  80. [88]

    Shaomei Wu, Jeffrey Wieland, Omid Farivar, and Julie Schiller. 2017. Automatic alt-text: Computer-generated image descriptions for blind users on a social network service. Inproceedings of the 2017 ACM conference on computer supported cooperative work and social computing. 1180–1192

    2017

  81. [89]

    Jingnan Xu, Xiaolin Zheng, and Weifeng Ding. 2012. Personalized recommenda- tion based on reviews and ratings alleviating the sparsity problem of collaborative filtering. In2012 IEEE Ninth International Conference on e-Business Engineering. IEEE, 9–16

    2012

  82. [90]

    Xianjun Yang, Yan Li, Xinlu Zhang, Haifeng Chen, and Wei Cheng. 2023. Explor- ing the limits of chatgpt for query or aspect-based text summarization.arXiv preprint arXiv:2302.08081(2023)

    2023 arXiv

  83. [91]

    Ali Yayli and Murat Bayram. 2012. e-WOM: the effects of online consumer reviews on purchasing decisions.International Journal of Internet Marketing and Advertising7 (2012), 51. https://api.semanticscholar.org/CorpusID:168100918

    2012

  84. [92]

    Jonathan Zong, Crystal Lee, Alan Lundgard, JiWoong Jang, Daniel Hajas, and Arvind Satyanarayan. 2022. Rich screen reader experiences for accessible data visualization. InComputer Graphics Forum, Vol. 41. Wiley Online Library, 15–27

    2022

  85. [93]

    Cristina Zuheros, Eugenio Martínez-Cámara, Enrique Herrera-Viedma, and Fran- cisco Herrera. 2021. Sentiment analysis based multi-person multi-criteria decision making methodology using natural language processing and deep learning for smarter decision aid. Case study of restau...

    2021

  86. [2017]

    In2017 Seventh International Conference on Information Science and Technology (ICIST)

    A new approach for modeling researcher community based on scientific article metadata using natural language processing. In2017 Seventh International Conference on Information Science and Technology (ICIST). IEEE, 386–390

  87. [2021]

    Automatic text summarization: A comprehensive survey.Expert systems with applications165 (2021), 113679

    2021

  88. [2024]

    InProceedings of the 2024 10th International Conference on e-Society, e-Learning and e-Technologies (ICSLT)

    Evaluation of ChatGPT, Gemini and Llama-2 for E-commerce Product Attribute Extraction. InProceedings of the 2024 10th International Conference on e-Society, e-Learning and e-Technologies (ICSLT). 43–48

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.