A Cross-Validation Study of Turkish Sentiment Analysis Datasets and Tools

Ali H\"urriyeto\u{g}lu; \c{S}evval \c{C}ak{\i}c{\i}; Dilara Karaduman; Mehmet Akif \c{C}{\i}rlan

REVIEW 4 major objections 5 minor 23 references

A Cross-Validation Study of Turkish Sentiment Analysis Datasets and Tools

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Turkish sentiment analysis tools are not portable across text types: the same model reaches .92 on one tweet set and .72 on a review set, while TSAM goes from .59 on tweets to .97 on long reviews.

desk verdict A genuinely useful map of Turkish sentiment resources, but Table 4's metrics are internally inconsistent and the headline claim is not yet supported. read the letter →

arxiv 2412.05964 v1 pith:JVCATP4D submitted 2024-12-08 cs.CL

Şevval Çak{i}c{i} , Dilara Karaduman , Mehmet Akif Ç{i}rlan , Ali Hürriyetoğlu This is my paper · ORCID

classification cs.CL

keywords sentimentanalysisTurkishcross-validationtaxonomydatasetsurveyXLM-TBERTurkBERTweet

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 tries to establish that Turkish sentiment analysis tools cannot be treated as a single reusable resource: the accuracy of a given tool depends strongly on the characteristics of the text it is applied to. To show this, the authors compiled 23 Turkish datasets used in 31 studies published between 2012 and 2022, labeled them with a taxonomy, and ran four state-of-the-art sentiment models on four recurrent datasets. The results show large swings, such as XLM-T at .92 on one tweet set but .72 on movie and product reviews, while TSAM jumps from .59 on tweets to .97 on long reviews. If true, this means reported single-dataset scores mislead practitioners, and dataset curation and preprocessing deserve as much attention as model choice.

What carries the argument

The argument is carried by two coordinated devices. The first is a taxonomy of sentiment analysis studies—opinion detection type, entity, target class, data specificity, granularity, and approach—used to label the 31 collected studies and, implicitly, to describe what makes each dataset what it is. The second is a cross-dataset evaluation matrix in which each model is run on each dataset under two protocols: the native multi-class protocol and a 'Both: Binary' protocol that excludes neutral predictions. The taxonomy organizes the field; the matrix produces the evidence of domain dependence, and the binary re-scoring is what makes the swings visible, since raw multi-class scores compress all models into a narrow band.

What would settle it

Re-run the four models on the four datasets without dropping any predictions and compare accuracy and F1 computed over every test instance. If the .92-versus-.72 and .59-versus-.97 gaps shrink or reorder when neutral predictions count as errors, the central claim about dataset dependence is weakened; if the gaps persist with all predictions counted, the claim is supported.

Watch

Extended reading notes

Core claim

The central claim is that the performance of the sentiment analysis tools significantly depends on the characteristics of the target text. The paper tests this by applying four state-of-the-art models—XLM-T, BERTurk fine-tuned on BounTi, TSAM, and TurkishBERTweet—to four recurrent Turkish datasets and tabulating the results. Under the binary re-scoring protocol, XLM-T reaches .92 accuracy on Twt but .72 on TMRD-TMPR, while TSAM scores .59 on Twt and .97 on Humir. The authors interpret such swings as evidence that accuracy is not a fixed property of a model; source platform, label source, text length, domain, and class balance all change what a tool can do. The intended contribution is a map of Turkish sentiment analysis over 2012–2022 plus an empirical warning against treating any single dataset score as a general measure.

Load-bearing premise

The load-bearing premise is that the 'Both: Binary' evaluation protocol is fair: it drops the model's neutral predictions before recomputing accuracy, and if those dropped instances are hard or mislabeled, the reported cross-dataset differences are artifacts of that exclusion.

Editorial extensions

If this is right

A model chosen for a Turkish sentiment task should be validated on text from the target platform and genre, not on a generic Turkish benchmark.
Leaderboard numbers from a single dataset cannot be compared across studies without controlling for dataset characteristics.
Classification-format mismatches, such as multi-class models applied to binary data, depress scores; resolving the mismatch by excluding neutral predictions can raise them substantially.
Modern deep models improve on classical baselines, with XLM-T's .92 on Twt compared with the .66 reported for multinomial naive Bayes, so older reported accuracies understate what is achievable.
Dataset construction choices—label source, balance, text length, and domain—are first-order variables in Turkish sentiment analysis, not minor details.

Reading between the lines

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

A standardized Turkish sentiment benchmark with fixed splits and mandatory reporting of both full-set and neutral-excluded metrics would make cross-study results comparable.
Dataset 'characteristics' could be turned into measurable features, such as platform, document length, label acquisition method, and class balance, and then used to predict how well a model will transfer to a new text type.
The same cross-dataset protocol could be applied to other morphologically rich languages, for example Arabic review-and-tweet pairs, to test whether this domain dependence is a Turkish-specific finding or a general property of sentiment tools.
The taxonomy labels could be extended with a 'source of labels' dimension—emoticon-based, rating-based, or human-annotated—because the paper's largest accuracy swings track exactly that distinction.

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.

Desk Editor's Note

read the letter

The thing worth knowing: the survey half of this paper is a genuine contribution, and the benchmark half is not yet trustworthy. The taxonomy-labeled inventory of 31 Turkish sentiment studies and 23 datasets, with the GitHub repository, is the most useful single map of this space I've seen. The zero-shot evaluation idea is reasonable.

The stress-test note is right and it lands on Table 4. For balanced binary datasets, macro-recall is definitionally equal to accuracy, and micro-recall is always equal to accuracy. The table reports XLM-T on Twt at .83 accuracy, .83 macro-recall, and .55 micro-recall. That micro-recall cannot be standard. In the 'Both: Binary' rows, after excluding neutral predictions, accuracy, micro-recall, micro-precision, and micro-F1 are all the same quantity on the remaining instances; the table lists .92 and .95 in the same row for XLM-T on Twt. Either the columns are mislabeled, the protocol is something other than what the text says, or the numbers are simply wrong. The paper gives no way to tell, and the headline claim about text characteristics leans on these numbers.

The next soft spot is interpretive. The four datasets differ in domain, source, size, class balance, and preprocessing all at once. That configuration can't support a clean causal claim that 'characteristics' drive performance. And the comparison to best-reported numbers from the original studies is not apples-to-apples, since training setups, splits, and evaluation protocols differ.

Credit where it's due: the SLR-style compilation, the taxonomy labels in Table 2, and the public repo are all useful. I would keep the survey, fix the benchmark, and re-run the metrics. This paper deserves a serious referee, but as written I would not cite the empirical numbers. If I were the editor, I'd send it out with a request for a corrected Table 4, a precise definition of the neutral-exclusion rule, and ideally per-instance predictions or runnable scripts that reproduce those exact values.

Referee Report

4 major / 5 minor

Summary. This paper compiles 31 Turkish sentiment analysis studies published between 2012 and 2022 and 23 associated datasets, labels the studies with the Rodrigues et al. taxonomy, and presents the resulting map in Tables 1–3. The authors then run four pre-trained sentiment models (XLM-T, BERTurk with BounTi, TSAM, and TurkishBERTweet) on four recurrent datasets (Twt, Turkmenoglu, TMRD-TMPR, Humir) and report accuracy, recall, precision, and F1 scores under three classification-format settings in Table 4. The central claim, stated in the abstract, is that the performance of sentiment analysis tools significantly depends on the characteristics of the target text. The paper also provides a GitHub repository with code and predictions.

Significance. If the reported benchmark numbers are correct and the evaluation protocol is fully specified, this study would be a useful resource for Turkish sentiment analysis: it collects and organizes a substantial set of datasets, provides a taxonomic overview of a decade of work, and offers a cross-dataset comparison using modern pre-trained models. The public availability of the code and data is a clear strength, as is the effort to compile datasets that were previously scattered or only available by request. However, as written, the central empirical claim rests on Table 4, whose evaluation protocol is not defined precisely enough to allow reproduction or verification. The paper does not currently support the strong claim of statistical or causal dependence on dataset characteristics, because the datasets differ on many axes simultaneously and no controlled comparison or significance test is provided.

major comments (4)

[Section 4, Table 4] The evaluation protocol for Table 4 is not defined precisely enough to be reproducible. For the 'Model: Multi, Dataset: Binary' rows, the paper states that scores are calculated without modifying the classification formats, but it does not state how macro-recall is computed when the gold set has no neutral instances. For balanced binary datasets such as Twt and Humir, standard macro-recall over the two gold classes equals accuracy, yet the table reports values like Accuracy .83 vs. Macro Recall .55 for XLM-T on Twt. These numbers are consistent only with a non-standard macro-recall that includes the absent neutral class with zero recall. The manuscript must specify the exact computation, including how classes with no gold instances are handled, and should provide the underlying contingency tables or per-instance predictions in the repository.
[Section 4, Table 4 note] The 'Both: Binary' setting is ambiguous. The text says scores are recalculated by excluding 'neutral' labels, but for datasets that are already binary (Twt, Humir, TMRD-TMPR) there are no neutral gold labels to exclude. The only possible interpretation is that instances where the model predicts 'neutral' are dropped before computing scores. This protocol is not stated explicitly, is not justified, and can inflate accuracy by allowing the model to abstain on hard cases. Moreover, it is unclear whether the micro metrics are computed on the same set of remaining instances as accuracy; for example, the XLM-T/Twt row reports Accuracy .92 and Micro Recall .95, which the reader cannot verify from the information given. Please state the exact decision rule (dropping or remapping) and report the number of instances dropped per model and dataset.
[Abstract; Section 4; Section 5] The claim that performance 'significantly depends on the characteristics of the target text' is not supported by a controlled comparison. The four datasets differ simultaneously in source (Twitter vs. review sites), domain (movies, products, hotels), annotation scheme (emoticon-based, star ratings, manual labels), size, and time period. No analysis isolates any single characteristic, and no statistical significance test is reported. The data show that performance varies across datasets, but 'significantly depends' implies a causal or at least statistically established relation. I recommend weakening the claim to 'varies across datasets' or adding a controlled analysis—for example, stratified comparisons or a significance test—before the claim can be maintained.
[Section 5, paragraph 6] The statement that 'this study has shown that XLM-T is more effective than BERTurk' in overcoming Turkish linguistic challenges is too strong. In Table 4, XLM-T is not consistently better than BERTurk: on Humir with 'Both: Binary', BERTurk achieves .86 accuracy and .85 for XLM-T, and on Turkmenoglu with 'Both: Binary' the two are close (.81 vs. .82). The conclusion should be qualified to reflect the dataset-dependent nature of the comparison, or supported with a statistical test across datasets.

minor comments (5)

[Table 3] The entry for Köksal and Özgür [2021] reports an accuracy of .0745 for BERTurk on BounTi; this appears to be a typo for .745 or .74 and should be corrected.
[Section 2.1] The text says 'Brief descriptions, contents, and classification types of five datasets are as follows' after stating that the study focuses on four datasets; this inconsistency should be fixed, and the relationship between TMRD, TMPR, and the combined name 'TMRD-TMPR' should be clarified.
[Title and Section 1] The phrase 'Cross-Validation' in the title is misleading: the paper does not perform k-fold or any standard cross-validation procedure. If the intended meaning is cross-dataset evaluation, that term should be used and defined.
[Table 4] The column headers of Table 4 are ambiguous, with 'Recall' and 'Macro' appearing without clear ordering. The authors should explicitly list the metric sequence in the caption (e.g., Accuracy, Macro Recall, Micro Recall, Macro Precision, Micro Precision, Macro F1, Micro F1) and ensure each reported value is assigned to the correct column.
[Section 4, paragraph 2] The sentence 'XLM-T achieved a remarkable accuracy and F1 score of .92 and .95, respectively' does not specify whether the F1 score is macro or micro; please clarify, since the table row contains both values.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the performance-dependence claim is an empirical comparison of external models and datasets, not a derivation from fitted inputs or self-defined target quantities.

full rationale

The paper's central claim is that sentiment analysis tool performance varies with target text characteristics. This claim is supported by running four pre-existing, externally trained models (XLM-T, BERTurk, TSAM, TurkishBERTweet) on four pre-existing Turkish datasets and reporting the measured metrics. No model parameter is fitted to the evaluation datasets, no target metric is defined in terms of the models' outputs, and no uniqueness theorem or load-bearing result is imported from the authors' own prior work. The only self-citation (Hürriyetoğlu et al. 2021) is contextual and does not justify the central empirical finding. The taxonomy labeling of 31 studies is a classification exercise, not a derivation of performance. Although the reviewer notes that Table 4's metric triples appear internally inconsistent under standard definitions (e.g., accuracy and macro-recall cannot both take those values on balanced binary sets), that is a correctness and reproducibility concern, not circularity: the claim does not reduce to its inputs by construction. The empirical basis may be unverifiable as written, but the reasoning chain is not circular. Therefore the appropriate circularity score is 0.

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

The central empirical claim rests on four assumptions: the Scopus query covers the field, dataset differences can be interpreted as text-characteristic effects, the binary conversion by excluding neutral predictions is valid, and off-the-shelf scores are comparable to original-study reported scores. No free parameters are fitted and no new entities are introduced.

assumptions (4)

domain assumption The Scopus query 'sentiment analysis' AND 'Turkish dataset' returns the full population of Turkish sentiment analysis studies from 2012 to 2022.
Section 2.1: the systematic map is based on this single query; relevant studies using different terminology or not indexed by Scopus may be missing, so the 10-year map may be incomplete.
domain assumption Performance differences across datasets can be attributed to dataset characteristics (source, domain, class balance, text length) rather than to confounded factors.
Sections 4 and 5: the main conclusion that performance significantly depends on the characteristics of the target text rests on comparing four datasets that differ along multiple dimensions simultaneously; no controlled experiment or statistical test isolates the effect.
ad hoc to paper Excluding model predictions of 'neutral' when converting to binary classification is a valid evaluation protocol.
Section 4, Table 4 note: 'Both: Binary ... excluding neutral labels'; this post-hoc exclusion is applied to model outputs and can inflate accuracy for binary datasets, and its validity is not discussed.
domain assumption Off-the-shelf, non-fine-tuned model scores are comparable to the best-reported accuracies from the original studies, despite different training protocols.
Sections 4 and 5 compare Table 4 scores with 'Best Reported Model' numbers in Table 3 without re-running original models under a common evaluation setup.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Cross-Validation Study of Turkish Sentiment Analysis Datasets and Tools." pith.science (2026). https://pith.science/paper/JVCATP4D

@misc{pith2026241205964,
  author       = {Pith},
  title        = {Pith review of: A Cross-Validation Study of Turkish Sentiment Analysis Datasets and Tools},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JVCATP4D}},
  note         = {Machine review of arXiv:2412.05964}
}

read the original abstract

In recent years, sentiment analysis has gained increasing significance, prompting researchers to explore datasets in various languages, including Turkish. However, the limited availability of Turkish datasets has led to their multifaceted usage in different studies, yielding diverse outcomes. To overcome this challenge, a rigorous review was conducted of research articles published between 2012 and 2022. 31 studies were listed, and 23 Turkish datasets obtained from publicly available sources and email requests used in these studies were collected. We labeled these 31 studies using a taxonomy. We provide a map of sentiment analysis datasets according to this taxonomy in Turkish over 10 years. Moreover, we run state-of-the-art sentiment analysis tools on these datasets and analyzed performance across popular Turkish sentiment datasets. We observed that the performance of the sentiment analysis tools significantly depends on the characteristics of the target text. Our study fosters a more nuanced understanding of sentiment analysis in the Turkish language.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 21 canonical work pages

[7]

Anke, and Jose Camacho-Collado s

Francesco Barbieri, Luis E. Anke, and Jose Camacho-Collado s. Xlm-t: Multilingual language models in twitter for sentiment analysis and beyond. arXiv preprint arXiv:2104.12250 ,

arXiv
[11]

Hayran and M

A. Hayran and M. Sert. Sentiment analysis on microblog data b ased on word embedding and fusion techniques. In 2017 25th Signal Processing and Communications Application s Conference (SIU), pages 1–4. IEEE,

work page 2017
[12]

doi: 10.1162/ dint a 00092

ISSN 2641-435X. doi: 10.1162/ dint a 00092. URL https://doi.org/10.1162/dint_a_00092. A. T. Kabakus. A novel covid-19 sentiment analysis in turkis h based on the combination of con- volutional neural network and bidirectional long-short te rm memory on twitter. Concurrency and Computation: Practice and Experience , 34(22):e6883,

work page doi:10.1162/dint_a_00092
[13]

Kitchenham and S

B. Kitchenham and S. Charters. Guidelines for performing sy stematic literature reviews in soft- ware engineering. Technical Report EBSE 2007-001, Keele Un iversity and Durham University Joint Report,

work page 2007
[15]

URL https://licensing.visualcapitalist.com/product/the- most-used-languages-on-the-internet/ . H. Mulki, H. Haddad, C. B. Ali, and ˙I. Babao˘ glu. Preprocessing impact on turkish sentiment analysis. In 2018 26th Signal Processing and Communications Application s Conference (SIU), pages 1–4. IEEE,

work page 2018
[17]

H. A. O˘ gul and A. G¨ uran. Imbalanced dataset problem in sent iment analysis. In 2019 4th International Conference on Computer Science and Engineering (UBMK), pages 313–317. IEEE,

work page 2019
[18]

Parlar and S

T. Parlar and S. A. ¨Ozel. A new feature selection method for sentiment analysis of turkish reviews. In 2016 International Symposium on INnovations in Intelligent Sy sTems and Ap- plications (INISTA), pages 1–6. IEEE,

work page 2016
[19]

doi: 10.1108/ ijwis-07-2017-0048. M. Rumelli, D. Akku¸ s, ¨O. Kart, and Z. I¸ sık. Sentiment analysis in turkish text wit h machine learning algorithms. In 2019 Innovations in intelligent systems and applications co nference (ASYU), pages 1–5. IEEE,

work page 2017

Show all 23 references

[20]

M. U. Salur and ˙I. Aydın. An annotated turkish aspect based sentiment analysis corpus for smart tourism. In 2021 Innovations in Intelligent Systems and Applications Conf erence (ASYU) , pages 1–6. IEEE,

2021
[22]

C ¸ oban, B.¨Ozyer, and G

¨O. C ¸ oban, B.¨Ozyer, and G. T. ¨Ozyer. Sentiment analysis for turkish twitter feeds. In 2015 23nd Signal Processing and Communications Applications Conf erence (SIU) , pages 2388–

2015
[1992]

doi: https://doi.org/10.1080/02699939208411068. B. Er¸ sahin, ¨O. Akta¸ s, D. Kılın¸ c, and M. Er¸ sahin. A hybrid sentiment an alysis method for turkish. Turkish Journal of Electrical Engineering and Computer Scie nces, 27(3):1780–1793,

doi
[2002]

doi: https: //doi.org/10.1016/S0376-6357(02)00078-5. E. Demirtas and M. Pechenizkiy. Cross-lingual polarity det ection with machine translation. In Proceedings of the Second International Workshop on Issues of Sentiment Discovery and Opinion Mining , pages 1–8. ACM,

doi
[2007]

K¨ oksal and A

A. K¨ oksal and A. ¨Ozg¨ ur. Twitter dataset and evaluation of transformers for turkish sentiment analysis. In 2021 29th Signal Processing and Communications Application s Conference (SIU), pages 1–4. IEEE,

2021
[2008]

U. U. Acikalin, B. Bardak, and M. Kutlu. Turkish sentiment an alysis using bert. In 2020 28th Signal Processing and Communications Applications Conferen ce (SIU) , pages 1–4. IEEE,

2020
[2014]

A. Ucan, B. Naderalvojoud, E. A. Sezer, and H. Sever. Sentiwo rdnet for new language: auto- matic translation approach. In 2016 12th International Conference on Signal-Image Technol- ogy & Internet-Based Systems (SITIS) , pages 308–315. IEEE,

2016
[2015]

C ¸ oban and G

¨O. C ¸ oban and G. T.¨Ozyer. Word2vec and clustering based twitter sentiment ana lysis. In 2018 International Conference on Artiﬁcial Intelligence and Data P rocessing (IDAP), pages 1–5. IEEE,

2018
[2016]

A. G. Eker, K. Eker, and N. Duru. Multi-class sentiment analy sis from turkish tweets with rnn. In 2021 6th International Conference on Computer Science and Engi neering (UBMK), pages 560–564. IEEE,

2021
[2018]

M. M. Mutlu and A. ¨Ozg¨ ur. A dataset and bert-based models for targeted sentim ent analysis on turkish texts. arXiv preprint arXiv:2205.04185 ,

arXiv
[2019]

Alnahas, F

13 D. Alnahas, F. A¸ sık, A. Kanturvardar, and A. M. ¨Ulkg¨ un. Opinion mining using lstm net- works ensemble for multi-class sentiment analysis in e-com merce. In 2022 3rd International Informatics and Software Engineering Conference (IISEC) , pages 1–6. IEEE,

2022
[2020]

S. E. Akin and T. Yildiz. Sentiment analysis through transfe r learning for turkish language. In 2019 IEEE International Symposium on INnovations in Intelligen t SysTems and Applications (INISTA), pages 1–6. IEEE,

2019
[2021]

Balli, M

C. Balli, M. S. Guzel, E. Bostanci, and A. Mishra. Sentimenta l analysis of twitter users from turkish content with natural language processing. Computational Intelligence and Neuro- science, 2022,

2022
[2022]

Alqaraleh

S. Alqaraleh. Novel turkish sentiment analysis system usin g convnet. 2021,

2021
[2024]

M. F. Amasyali, H. Task¨ opr¨ u, and K. C ¸ aliskan. Words, mean ings, characters in sentiment analysis. In 2018 Innovations in Intelligent Systems and Applications Conf erence (ASYU) , pages 1–6. IEEE,

2018

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [7]

Anke, and Jose Camacho-Collado s

Francesco Barbieri, Luis E. Anke, and Jose Camacho-Collado s. Xlm-t: Multilingual language models in twitter for sentiment analysis and beyond. arXiv preprint arXiv:2104.12250 ,

arXiv

[2] [11]

Hayran and M

A. Hayran and M. Sert. Sentiment analysis on microblog data b ased on word embedding and fusion techniques. In 2017 25th Signal Processing and Communications Application s Conference (SIU), pages 1–4. IEEE,

work page 2017

[3] [12]

doi: 10.1162/ dint a 00092

ISSN 2641-435X. doi: 10.1162/ dint a 00092. URL https://doi.org/10.1162/dint_a_00092. A. T. Kabakus. A novel covid-19 sentiment analysis in turkis h based on the combination of con- volutional neural network and bidirectional long-short te rm memory on twitter. Concurrency and Computation: Practice and Experience , 34(22):e6883,

work page doi:10.1162/dint_a_00092

[4] [13]

Kitchenham and S

B. Kitchenham and S. Charters. Guidelines for performing sy stematic literature reviews in soft- ware engineering. Technical Report EBSE 2007-001, Keele Un iversity and Durham University Joint Report,

work page 2007

[5] [15]

URL https://licensing.visualcapitalist.com/product/the- most-used-languages-on-the-internet/ . H. Mulki, H. Haddad, C. B. Ali, and ˙I. Babao˘ glu. Preprocessing impact on turkish sentiment analysis. In 2018 26th Signal Processing and Communications Application s Conference (SIU), pages 1–4. IEEE,

work page 2018

[6] [17]

H. A. O˘ gul and A. G¨ uran. Imbalanced dataset problem in sent iment analysis. In 2019 4th International Conference on Computer Science and Engineering (UBMK), pages 313–317. IEEE,

work page 2019

[7] [18]

Parlar and S

T. Parlar and S. A. ¨Ozel. A new feature selection method for sentiment analysis of turkish reviews. In 2016 International Symposium on INnovations in Intelligent Sy sTems and Ap- plications (INISTA), pages 1–6. IEEE,

work page 2016

[8] [19]

doi: 10.1108/ ijwis-07-2017-0048. M. Rumelli, D. Akku¸ s, ¨O. Kart, and Z. I¸ sık. Sentiment analysis in turkish text wit h machine learning algorithms. In 2019 Innovations in intelligent systems and applications co nference (ASYU), pages 1–5. IEEE,

work page 2017