Pith. sign in

REVIEW 4 major objections 6 minor 1 cited by

AdParaphrase: Paraphrase Dataset for Analyzing Linguistic Features toward Generating Attractive Ad Texts

T0 review · 4 major / 6 minor · reviewed 2026-08-08 · deepseek-v4-flash

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

Pith's one-line read Paraphrased ads are judged more attractive when they are more fluent, longer, more noun-heavy, and use bracket symbols.

desk verdict Useful new resource, but the feature-attractiveness findings rest on a post-hoc high-agreement subset and unadjusted chi-squares; the dataset deserves review, the statistical claims need rework. read the letter →

arxiv 2502.04674 v2 pith:L7AZ5Q7N submitted 2025-02-07 cs.CL cs.AI

Soichiro Murakami , Peinan Zhang , Hidetaka Kamigaito , Hiroya Takamura , Manabu Okumura This is my paper · ORCID
classification cs.CLcs.AI
keywords adtextgenerationparaphrasedatasethumanpreferenceslinguisticfeaturesfew-shotin-contextlearningJapanesetextsattractivenessevaluationchi-squareanalysis
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

This paper introduces AdParaphrase, a dataset of 725 pairs of Japanese ad texts that mean the same thing but differ in wording and style, each pair judged for attractiveness by ten people. It uses these matched pairs to isolate style from content, then tests 24 linguistic features against majority preference. The paper's central finding is that the ad text most people prefer tends to be more fluent (lower perplexity), longer in characters, richer in nouns and noun phrases, and likelier to contain Japanese bracket symbols, with all four associations significant at $p<0.01$. It then turns the finding into a prompt for GPT-4 that includes positive and negative before/after examples; that prompted model reaches a 36.6% attractive-success rate and 30.8% overall success in human evaluation, above the 23.2% attractiveness of paraphrases written by human advertising experts. The broader point is that content-independent stylistic preferences in ads can be measured with paired data and then used to make machine rewriting attractive.

What carries the argument

The central mechanism is the paired-paraphrase design: collecting ad-text pairs that are semantically equivalent, so that any systematic preference between the two members can only be attributed to wording and style. On those pairs, a chi-square test of independence tests whether the text preferred by the majority of judges is also the one scoring higher on each of 24 linguistic features. The generation side uses in-context learning: a GPT-4 prompt that states the four findings as 'tips' and supplies both positive demonstrations (where a rewrite was preferred) and negative demonstrations (where judges did not converge), which teaches the model both what attractiveness looks like and how aggressively to edit.

What would settle it

Re-run the feature-versus-preference comparison on all 725 pairs using the continuous share of judges preferring each text instead of the majority label on only the 316 high-agreement pairs; if lower perplexity, longer length, noun density, and bracket use no longer predict preference, the central finding fails.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that human attractiveness judgments for paraphrased ad texts are not arbitrary: in the 316 of 725 pairs where at least seven of ten judges agreed, the majority-chosen text has significantly lower perplexity, significantly more characters, significantly more nouns and noun phrases, and significantly more bracket usage (all $p<0.01$ by chi-square test). The paper treats these as the linguistic signature of an attractive Japanese ad text. It further claims that encoding this signature into a GPT-4 instruction prompt, together with twenty positive examples (input less-preferred to output more-preferred) and twenty negative examples (pairs without consensus), yields ad rewrites that human judges find attractive 36.6% of the time and paraphrase-acceptable-plus-attractive 30.8% of the time, outperforming zero-shot GPT-4, few-shot variants without the findings, and human expert paraphrases (23.2% attractive).

Load-bearing premise

The analysis assumes the 316 pairs where at least seven judges agreed are representative, so the style preferences found on that subset would also hold for the 409 pairs where judges disagreed, and features like length or brackets do not themselves influence how much judges agree.

Editorial extensions

If this is right

  • Ad copywriters can act on four concrete style levers: write closer to natural fluency, use the full character budget, favor nouns and noun phrases, and mark key phrases with brackets.
  • LLM-based ad rewriting can exceed the attractiveness of human-expert paraphrases when prompted with preference-derived findings plus positive and negative examples, at least for Japanese search ads.
  • Ad texts preferred by seven or more judges align with predicted CTR in 54.7% of cases, so style improvements that win human votes can translate into modest click-performance gains.
  • The public dataset lets other researchers replay the preference analysis and test new features or rewriting strategies without access to proprietary ad log data.
  • Negative few-shot examples, drawn from pairs without clear consensus, appear to calibrate the amount of editing a model performs; models given only positive examples made smaller lexical changes and improved less.

Reading between the lines

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

  • The contrastive paraphrase design could be exported to other content types, such as product descriptions, headlines, or notifications, to separate style-driven preference from content-driven preference in the same way.
  • The bracket effect is plausibly tied to the visual landscape of Japanese search ads and the 15-character headline limit, so the same feature may not transfer to banner ads, long-form copy, or other languages; testing this would require a multilingual paired-preference dataset.
  • The 316-pair subset masks disagreement: if long or bracketed texts polarize judges, the feature signature could partly describe texts that produce consensus rather than texts that are broadly attractive.
  • A direct online A/B test with more ad groups would tell whether the best prompted model reliably lifts clicks and impressions; the paper's own three-group test improved only one of three ad groups.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper introduces AdParaphrase, a Japanese ad-text paraphrase dataset with human attractiveness judgments. It constructs 725 paraphrase pairs from existing datasets and human/LLM-generated paraphrases, collects ten preference judgments per pair, analyzes 24 linguistic features via chi-square tests on the 316 pairs with at least 7/10 judge agreement, and reports that preferred texts exhibit lower perplexity, longer character counts, more nouns and noun phrases, and greater bracket use. It then incorporates these findings into GPT-4 prompts with positive and negative few-shot examples, reporting human-evaluated improvements in paraphrase identification and attractiveness, and includes an analysis of alignment with predicted CTR and a small online A/B test.

Significance. If the feature findings are valid, AdParaphrase is a valuable public resource: the paraphrase-pair design controls for semantic content, the preference judgments use ten judges per pair, and the dataset, prompts, and analysis code are publicly released on GitHub. The generation experiment is concrete and falsifiable, with GPT-4-fewshot-findings-both reaching 36.6% attractiveness versus 23.2% for human expert paraphrases. However, the central statistical claim is currently undermined by a post-hoc subset selection and by the lack of multiple-comparison correction and significance testing for key generation comparisons, so the findings should be treated as promising but not yet established.

major comments (4)
  1. [§4.1.2–§4.1.3, Table 3] The chi-square analysis is restricted to the 316 of 725 pairs in which at least 7 of 10 judges agreed on the more attractive text. Because this filter is applied after observing the preference outcome, it is not a neutral design choice: if a linguistic feature (e.g., bracket use or length) polarizes judges, the probability of reaching 7/10 agreement is correlated with that feature, so associations in the selected subset need not reflect the full population. The paper offers no comparison of feature distributions across agreement levels and no analysis of the 409 excluded pairs. Since every reported feature-preference conclusion (lower PPL, longer character counts, more nouns/noun phrases, more brackets) derives from this subset, the central claim of the abstract and §4.1.3 is not demonstrated for the dataset as a whole. Please repeat the analysis on all 725 pairs (e.g., using a mixed-effects logistic regression or a continuous measure of preference strength) or provide a formal argument that agreement is independent of the tested features.
  2. [Table 3] Twenty-four chi-square tests are reported without any multiple-comparison correction. With a threshold of p<0.01, roughly 0.24 false positives are expected, and several p-values are close to 0.01 (word-length p=0.02, kanji p=0.02). Many features are also correlated (e.g., character length, noun count, noun phrases, and perplexity), which further complicates interpretation. Please apply a multiple-comparison correction (e.g., Benjamini-Hochberg or Bonferroni) and report adjusted p-values, or explicitly justify why unadjusted tests are appropriate.
  3. [§4.2.3, Table 4, Appendix F] The claim that findings 'significantly improved performance' is not supported by a significance test: the only Fisher's exact test reported compares GPT-4-fewshot-both with GPT-4-fewshot-pos, not the findings-versus-no-findings comparisons, and the experiment is a single run (Appendix F states 'We report the results of a single experiment'). The success rates in Table 4 are therefore point estimates with unknown variance, and the superiority over the HUMAN baseline (36.6% vs. 23.2%) is not shown to be statistically reliable. Please provide confidence intervals, repeated runs, or significance tests for the key comparisons, ideally including GPT-4-zeroshot-findings versus GPT-4-zeroshot and GPT-4-fewshot-findings-both versus HUMAN.
  4. [§4.2, Fig. 9, §4.2.3] The generation experiment has a circularity component: the findings embedded in the prompt are learned from the same type of human-preference judgments used to evaluate the generated texts, and the concluding sentence of §4.2.3 states that the results 'suggest that the linguistic features identified through the linguistic feature analysis (§4.1) contribute to enhancing the attractiveness of ad texts.' This is an application rather than an independent confirmation of a causal effect. The paper should explicitly acknowledge this and soften the causal wording, presenting the generation results as demonstrating that prompt-engineering with these features can steer an LLM on the same evaluation metric.
minor comments (6)
  1. [§2.1] The paper calls one source dataset 'Ad Similarity' and cites Zhang et al. (2024), but the reference list entry for that citation is titled 'AdTEC: A unified benchmark for evaluating text quality in search engine advertising.' Please clarify the relationship between Ad Similarity and AdTEC and ensure the citation name matches the dataset name.
  2. [§3.2] The sentence 'we randomized the order of ad text pairs before presenting them to the judges' is ambiguous: it should state whether the order of the two texts within each pair was randomized, which is what mitigates position bias.
  3. [Figure 3] The x-axis is labeled 'maximum number of votes between ad text pair'; the text explains that zero means all judges skipped, but it does not explain how ties other than 'all skip' are represented (e.g., 5–5 splits). Please add a clarifying sentence.
  4. [§4.2.3] The text reports average Jaccard similarities of 0.560 and 0.602 for positive and negative examples, while Table 5 reports Jaccard between input and generated text; make clear that these are different quantities and that the table reports generation similarity, not example similarity.
  5. [§5.1, Table 7] The pCTR alignment rates (54.7% in the high-agreement group and 44.5% in the low-agreement group) are compared informally; since 54.7% is not far from chance, please add a statistical test (e.g., a binomial test) and note that pCTR is a model prediction rather than observed click data.
  6. [Appendix F] The statement 'We report the results of a single experiment' should appear in the main text or at least in the main experimental section, not only in the appendix, so that readers are not misled about the precision of Table 4.

Circularity Check

1 steps flagged · score 3.0 of 10

The generation experiment re-encodes the fitted preference signal, but the feature analysis and dataset contribution are independently grounded.

  1. fitted input called prediction [§4.1.2, §4.2.1–§4.2.3 (analysis subset, prompt construction, success definition)]
    "To ensure the reliability of our analysis, we used the 316 (out of 725) pairs of ad texts, in which the choices of more than seven human judges for attractiveness were consistent, as this indicated fair agreement... For attractiveness, we defined the success rate as the proportion of paraphrased texts found to be more attractive by at least seven out of ten judges."

    The feature analysis and the generation success metric use the same operational definition of 'attractive' (majority preference of ≥7/10 judges). The findings from that analysis—brackets, full character limit, fluency—are then written into the generation prompt, and positive few-shot examples are selected using the same ≥7/10 agreement criterion. The generation experiment therefore scores the model with the same ruler that was used to select the features and demonstrations; Table 6 mainly verifies that GPT-4 followed the prompt, and Table 4 re-encodes the learned preference. This is an application of the findings rather than an independent confirmation that these features cause attractiveness. New judges and held-out inputs prevent it from being fully circular.

full rationale

The core chi-square analysis in §4.1 is self-contained: the associations between majority preference and linguistic features are estimated directly from the collected human judgments, not derived from the hypothesis or from a fitted parameter. The post-hoc restriction to the 316 high-agreement pairs is a selection-bias and generalizability concern, not a circularity, so it does not raise the circularity score by itself. The generation experiment introduces partial circularity because the prompt incorporates the previously discovered features and the evaluation uses the same majority-preference threshold that defined those discoveries; observing that the prompt-following model is preferred does not independently validate the causal role of the features. The self-citations are minor and not load-bearing: the pCTR-proxy justification via Mita et al. (2024) is checked directly in §5.1, and the dataset construction citations are standard source attribution. Overall, the paper's dataset and descriptive findings have independent content; only the generation 'demonstration' partly re-encodes its inputs.

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

The paper does not introduce a formal model or derivation, so no mathematical axioms are needed. The free parameters are hand-chosen thresholds and instruction settings that directly shape the dataset and the analysis. The domain assumptions are the empirical premises that connect the measured human preferences to real advertising performance.

free parameters (4)
  • Ad Similarity similarity threshold = 4 on a five-point scale
    Pairs with average similarity >=4 are chosen as candidates; this threshold shapes the dataset and is not derived from data.
  • High-agreement analysis threshold = 7 of 10 judges
    Feature analysis is restricted to pairs with >=7 agreeing judges; the choice affects all chi-square results.
  • Paraphrase length cap = 15 full-width characters
    Instructed to human experts and LLMs in dataset construction and generation, may drive the longer-text preference.
  • Few-shot example count = 20 positive + 20 negative
    Chosen without ablation; the generation results depend on this number.
assumptions (4)
  • domain assumption Human attractiveness ratings are a valid proxy for ad performance (pCTR)
    Cited from prior work in §3 and used to justify preference collection; the paper's own alignment is only 54.7% for high-agreement pairs (§5.1).
  • domain assumption Majority-voted paraphrase labels are correct
    Used to define the 725-pair evaluation set; Fleiss' kappa of 0.462 (§2.3) indicates moderate agreement, so some 'paraphrases' may differ in content.
  • domain assumption GPT-2 perplexity measures fluency of Japanese ad text
    Used as the fluency feature in §4.1; no validation on ad domain, and the model is trained on web corpora.
  • domain assumption The 133 CAMERA source texts are representative for evaluating ad text refinement
    Used as test inputs in §4.2.2; no evidence that they cover the diversity of real ad text.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AdParaphrase: Paraphrase Dataset for Analyzing Linguistic Features toward Generating Attractive Ad Texts." pith.science (2026). https://pith.science/paper/L7AZ5Q7N

@misc{pith2026250204674,
  author       = {Pith},
  title        = {Pith review of: AdParaphrase: Paraphrase Dataset for Analyzing Linguistic Features toward Generating Attractive Ad Texts},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/L7AZ5Q7N}},
  note         = {Machine review of arXiv:2502.04674}
}
read the original abstract

Effective linguistic choices that attract potential customers play crucial roles in advertising success. This study aims to explore the linguistic features of ad texts that influence human preferences. Although the creation of attractive ad texts is an active area of research, progress in understanding the specific linguistic features that affect attractiveness is hindered by several obstacles. First, human preferences are complex and influenced by multiple factors, including their content, such as brand names, and their linguistic styles, making analysis challenging. Second, publicly available ad text datasets that include human preferences are lacking, such as ad performance metrics and human feedback, which reflect people's interests. To address these problems, we present AdParaphrase, a paraphrase dataset that contains human preferences for pairs of ad texts that are semantically equivalent but differ in terms of wording and style. This dataset allows for preference analysis that focuses on the differences in linguistic features. Our analysis revealed that ad texts preferred by human judges have higher fluency, longer length, more nouns, and use of bracket symbols. Furthermore, we demonstrate that an ad text-generation model that considers these findings significantly improves the attractiveness of a given text. The dataset is publicly available at: https://github.com/CyberAgentAILab/AdParaphrase.

Figures

Figures reproduced from arXiv: 2502.04674 by the authors.

Figure 1
Figure 1. Overview of human preference judgments for [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Distribution of Jaccard similarity for para [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Distribution of maximum number of votes between ad text pair in attractiveness evaluation. guidelines of Wang et al. (2021), offering examples of evaluation perspectives such as catchy, mem￾orable, and easy-to-read. (3) To ensure the an￾notation quality across crowdsourced workers of varying skill levels, we incorporated dummy ques￾tions with clear answers. These dummy questions consisted of pairs of identical ad te… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Instruction for paraphrase creation presented [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Prompt for paraphrase creation presented to [PITH_FULL_IMAGE:figures/full_fig_p012_5.png]
Figure 8
Figure 8. Figure 8: User interface for attractiveness evaluation. [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: Prompt for ad text generation experiment. For [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. AdParaphrase v2.0: Generating Attractive Ad Texts Using a Preference-Annotated Paraphrase Dataset

    cs.CL 2025-05 conditional novelty 5.0 of 10

    AdParaphrase v2.0 provides 16,460 preference-annotated Japanese ad paraphrase pairs, revealing linguistic correlates of attractiveness and supporting DPO-based rewriting, though labels are noisy and online evidence is...

Reference graph

Works this paper leans on

34 extracted references · 12 canonical work pages · cited by 1 Pith paper

  1. [1]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRING...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Kevin Bartz, Cory Barr, and Adil Aijaz. 2008. https://doi.org/10.1145/1386790.1386792 Natural language generation for sponsored-search advertisements . In Proceedings of the 9th ACM Conference on Electronic Commerce, pages 1--9

    work page arXiv 2008

  4. [4]

    Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, Sandhini Agarwal, Ariel Herbert-Voss, Gretchen Krueger, Tom Henighan, Rewon Child, Aditya Ramesh, Daniel Ziegler, Jeffrey Wu, Clemens Winter, Chris Hesse, Mark Chen, Eric Sigler, Mateusz Litwin, Scott Gr...

    work page 2020

  5. [5]

    Jan Cegin, Jakub Simko, and Peter Brusilovsky. 2023. https://doi.org/10.18653/v1/2023.emnlp-main.117 C hat GPT to replace crowdsourcing of paraphrases for intent classification: Higher diversity and comparable model robustness . In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 1889--1905

    doi:10.18653/v1/2023.emnlp-main.117 2023

  6. [6]

    Patrick Fernandes, Aman Madaan, Emmy Liu, António Farinhas, Pedro Henrique Martins, Amanda Bertsch, José G. C. de Souza, Shuyan Zhou, Tongshuang Wu, Graham Neubig, and André F. T. Martins. 2023. https://doi.org/10.1162/tacl_a_00626 Bridging the Gap: A Survey on Integrating (Human) Feedback for Natural Language Generation . Transactions of the Association ...

    doi:10.1162/tacl_a_00626 2023

  7. [7]

    Fleiss et al

    J.L. Fleiss et al. 1971. Measuring nominal scale agreement among many raters . Psychological Bulletin, 76(5):378--382

    1971

  8. [8]

    Yebowen Hu, Kaiqiang Song, Sangwoo Cho, Xiaoyang Wang, Hassan Foroosh, and Fei Liu. 2023. https://doi.org/10.18653/v1/2023.emnlp-main.519 D ecipher P ref: Analyzing influential factors in human preference judgments via GPT -4 . In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 8344--8357

    work page doi:10.18653/v1/2023.emnlp-main.519 2023

Show all 34 references
  1. [9]

    Weston Hughes, Keng-hao Chang, and Ruofei Zhang

    J. Weston Hughes, Keng-hao Chang, and Ruofei Zhang. 2019. https://doi.org/10.1145/3292500.3330754 Generating better search engine text advertisements with deep reinforcement learning . In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Da...

    2019

  2. [10]

    Tomoyuki Kajiwara, Chenhui Chu, Noriko Takemura, Yuta Nakashima, and Hajime Nagahara. 2021. https://doi.org/10.18653/v1/2021.naacl-main.169 WRIME : A new dataset for emotional intensity estimation with subjective and objective annotations . In Proceedings of the 2021 Conferenc...

    2021 doi

  3. [11]

    Hidetaka Kamigaito, Peinan Zhang, Hiroya Takamura, and Manabu Okumura. 2021. https://aclanthology.org/2021.naacl-industry.32/ An empirical study of generating texts for search engine advertising . In Proceedings of the 2021 Conference of the North American Chapter of the Assoc...

    2021

  4. [12]

    Hannah Kirk, Andrew Bean, Bertie Vidgen, Paul Rottger, and Scott Hale. 2023. https://doi.org/10.18653/v1/2023.emnlp-main.148 The past, present and better future of feedback learning in large language models for subjective human preferences and values . In Proceedings of the 20...

    2023 doi

  5. [13]

    Richard Landis and Gary G

    J. Richard Landis and Gary G. Koch. 1977. http://www.jstor.org/stable/2529310 The measurement of observer agreement for categorical data . Biometrics, 33(1):159--174

    1977

  6. [14]

    Kikuo Maekawa, Makoto Yamazaki, Takehiko Maruyama, Masaya Yamaguchi, Hideki Ogura, Wakako Kashino, Toshinobu Ogiso, Hanae Koiso, and Yasuharu Den. 2010. http://www.lrec-conf.org/proceedings/lrec2010/pdf/99_Paper.pdf Design, compilation, and preliminary analyses of B alanced C ...

    2010

  7. [15]

    Shaunak Mishra, Manisha Verma, Yichao Zhou, Kapil Thadani, and Wei Wang. 2020. https://doi.org/10.1145/3340531.3412720 Learning to create better ads: Generation and ranking approaches for ad creative refinement . In Proceedings of the 29th ACM International Conference on Infor...

    2020

  8. [16]

    Masato Mita, Soichiro Murakami, Akihiko Kato, and Peinan Zhang. 2024. https://arxiv.org/abs/2309.12030 Striking gold in advertising: Standardization and exploration of ad text generation . Preprint, arXiv:2309.12030

    2024 arXiv

  9. [17]

    Soichiro Murakami, Sho Hoshino, and Peinan Zhang. 2023. https://arxiv.org/abs/2306.12719 Natural language generation for advertising: A survey . Preprint, arXiv:2306.12719

    2023 arXiv

  10. [18]

    Soichiro Murakami, Peinan Zhang, Sho Hoshino, Hidetaka Kamigaito, Hiroya Takamura, and Manabu Okumura. 2022. https://doi.org/10.18653/v1/2022.naacl-industry.9 Aspect-based analysis of advertising appeals for search engine advertising . In Proceedings of the 2022 Conference of ...

    2022 doi

  11. [19]

    OpenAI. 2024. https://arxiv.org/abs/2303.08774 GPT-4 technical report . Preprint, arXiv:2303.08774

    2024 arXiv

  12. [20]

    Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul F Christiano, Jan Leike, a...

    2022

  13. [21]

    Reid Pryzant, Sugato Basu, and Kazoo Sone. 2018. https://doi.org/10.18653/v1/W18-5415 Interpretable neural architectures for attributing an ad ' s performance to its writing style . In Proceedings of the 2018 EMNLP Workshop B lackbox NLP : Analyzing and Interpreting Neural Net...

    2018 doi

  14. [22]

    Alec Radford, Jeff Wu, Rewon Child, David Luan, Dario Amodei, and Ilya Sutskever. 2019. https://openai.com/blog/better-language-models/ Language models are unsupervised multitask learners . OpenAI Blog

    2019

  15. [23]

    Satoshi Sato, Suguru Matsuyoshi, and Yohsuke Kondoh. 2008. http://www.lrec-conf.org/proceedings/lrec2008/pdf/165_paper.pdf Automatic assessment of J apanese text readability based on a textbook corpus . In Proceedings of the Sixth International Conference on Language Resources...

    2008

  16. [24]

    Lingfeng Shen, Lemao Liu, Haiyun Jiang, and Shuming Shi. 2022. https://doi.org/10.18653/v1/2022.emnlp-main.208 On the evaluation metrics for paraphrase generation . In Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing, pages 3178--3190

    2022 doi

  17. [25]

    Kazuma Takaoka, Sorami Hisamoto, Noriko Kawahara, Miho Sakamoto, Yoshitaka Uchida, and Yuji Matsumoto. 2018. Sudachi: a japanese tokenizer for business. In Proceedings of the Eleventh International Conference on Language Resources and Evaluation

    2018

  18. [26]

    Stamatina Thomaidou, Ismini Lourentzou, Panagiotis Katsivelis-Perakis, and Michalis Vazirgiannis. 2013. https://doi.org/10.1145/2505515.2507876 Automated snippet generation for online advertising . In Proceedings of the 22nd ACM International Conference on Information and Know...

    2013

  19. [27]

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, W...

    2023 arXiv

  20. [28]

    Hsueh-Cheng Wang and Marc Pomplun. 2012. https://doi.org/10.1167/12.6.26 The attraction of visual attention to texts in real-world scenes . Journal of Vision, 12(6):26--26

    2012 doi

  21. [29]

    Xiting Wang, Xinwei Gu, Jie Cao, Zihua Zhao, Yulan Yan, Bhuvan Middha, and Xing Xie. 2021. https://doi.org/10.1145/3447548.3467105 Reinforcing pretrained models for generating attractive text advertisements . In Proceedings of the 27th ACM SIGKDD International Conference on Kn...

    2021

  22. [30]

    Penghui Wei, Xuanhua Yang, Shaoguo Liu, Liang Wang, and Bo Zheng. 2022. https://aclanthology.org/2022.naacl-industry.2/ CREATER : CTR -driven advertising text generation with controlled pre-training and contrastive fine-tuning . In Proceedings of the 2022 Conference of the Nor...

    2022

  23. [31]

    Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, and Yuji Matsumoto. 2020. https://doi.org/10.18653/v1/2020.emnlp-main.523 LUKE : Deep contextualized entity representations with entity-aware self-attention . In Proceedings of the 2020 Conference on Empirical Methods ...

    2020 doi

  24. [32]

    Brit Youngmann, Elad Yom-Tov, Ran Gilad-Bachrach, and Danny Karmon. 2020. https://doi.org/10.1145/3366423.3380211 The automated copywriter: Algorithmic rephrasing of health-related advertisements to improve their performance . In Proceedings of The Web Conference 2020, pages 1...

    2020

  25. [33]

    Yuan Yuan, Fengli Xu, Hancheng Cao, Guozhen Zhang, Pan Hui, Yong Li, and Depeng Jin. 2023. https://doi.org/10.1109/TKDE.2021.3110724 Persuade to click: Context-aware persuasion model for online textual advertisement . IEEE Transactions on Knowledge and Data Engineering, 35(2):...

    2023

  26. [34]

    Peinan Zhang, Yusuke Sakai, Masato Mita, Hiroki Ouchi, and Taro Watanabe. 2024. https://arxiv.org/abs/2408.05906 A d TEC : A unified benchmark for evaluating text quality in search engine advertising . Preprint, arXiv:2408.05906

    2024 arXiv

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.