Desktop versionMobile Version

EVALITA Evaluation of NLP and Speech Tools for Italian - December 17th, 2020

 | 
Valerio Basile
, 
Danilo Croce
, 
Maria Maro
, 
et al.

AMI: Automatic Misogyny Identification

AMI @ EVALITA2020: Automatic Misogyny Identification

Elisabetta Fersini, Debora Nozza und Paolo Rosso

Zusammenfassung

Automatic Misogyny Identification (AMI) is a shared task proposed at the Evalita 2020 evaluation campaign. The AMI challenge, based on Italian tweets, is organized into two subtasks: (1) Subtask A about misogyny and aggressiveness identification and (2) Subtask B about the fairness of the model. At the end of the evaluation phase, we received a total of 20 runs for Subtask A and 11 runs for Subtask B, submitted by 8 teams. In this paper, we present an overview of the AMI shared task, the datasets, the evaluation methodology, the results obtained by the participants and a discussion about the methodology adopted by the teams. Finally, we draw some conclusions and discuss future work.

Volltext

The work of the last author was partially funded by the Spanish MICINN under the research project MISMISFAKEnHATE on MISinformation and MIScommunication in social media: FAKE news and HATE speech (PGC2018-096212-B-C31) and by the COST Action 17124 DigForAsp supported by the European Cooperation in Science and Technology.

1. Introduction

1The expressions of people about thoughts, emotions, and feelings by means of posts in social media have been widely spread. Women have a strong presence in these online environments: 75% of females use social media multiple times per day compared to 64% of males. While new opportunities emerged for women to express themselves, systematic inequality and discrimination take place in the form of offensive content against the female gender. These manifestations of misogyny, usually provided by a man to a woman for dominating or using a sort of power against the female gender, is a relevant social problem that has been addressed in the scientific literature during the last few years. Recent investigations studied how the misogyny phenomenon takes place, for example as unjustified slurring or as stereotyping of the role/body of a woman (i.e., the hashtag #getbacktokitchen), as described in the book by Poland (Poland 2016). Preliminary research work was conducted in (Hewitt et al., 2016) as the first attempt of manual classification of misogynous tweets, while automatic misogyny identification in social media has been firstly investigated in (Anzovino, Fersini, and Rosso 2018). Since 2018, several initiatives have been dedicated as a call-to-action to stop hate against women both from a machine learning and computational linguistics points of view, such as AMI@Evalita 2018 (Fersini, Nozza, and Rosso 2018), AMI@IberEval2018 (Fersini, Rosso, and Anzovino 2018) and HatEval@SemEval2019 (Basile et al. 2019). Several relevant research directions have been investigated for addressing the misogyny identification challenge, among which approaches focused on effective text representation (Bakarov 2018; Basile and Rubagotti 2018), machine learning models (Buscaldi 2018; Ahluwalia et al. 2018) and domain-specific lexical resources (Pamungkas et al. 2018; Frenda et al. 2018).

2During the AMI shared task organized at the Evalita 2020 evaluation campaign (Basile et al. 2020), the focus is not only on misogyny identification but also on aggressiveness recognition, as well as to the definition of models able to guarantee fair predictions.

2. Task Description

3The AMI shared task, which is a re-run of a previous challenge at Evalita 2018, proposes the automatic identification of misogynous content in the Italian language on Twitter. More specifically, it is organized according to two main subtasks:

Subtask A - Misogyny & Aggressive Behaviour Identification

4A system must recognize if a text is misogynous or not, and in case of misogyny, if it expresses an aggressive attitude. In order to provide an annotated corpus for Subtask A, the following definitions have been adopted to label the collected dataset:

    • Misogynous: a text that expresses hating towards women in particular (in the form of insulting, sexual harassment, threats of violence, stereotype, objectification, and negation of male responsibility).

    • Not Misogynous: a text that does not express any form of hate towards women.

    • Aggressive: a message is considered aggressive if it (implicitly or explicitly) presents, incites, threatens, implies, suggests, or alludes to:
      • attitudes, violent actions, hostility or commission of offenses against women;
      • social isolation towards women for physical or psychological characteristics;
      • justify or legitimize an aggressive action against women.

    • Not Aggressive: If none of the previous conditions hold.

Subtask B - Unbiased Misogyny Identification

5A system must discriminate misogynistic contents from the non-misogynistic ones, while guaranteeing the fairness of the model (in terms of unintended bias) on a synthetic dataset (Nozza, Volpetti, and Fersini 2019). To this purpose Subtask B has the goal of measuring the attitude of a model to be fair when processing sentences containing specific identity terms that likely conveyed misogyny in the training data, e.g. “girlfriend" and “wife".

3. Training and Testing Data

6The data provided to the participants for the AMI shared task comprises a raw dataset and a synthetic dataset for measuring bias. Each dataset is distinguished in Training Set and Test Set.

3.1 Raw dataset

7The raw dataset is a balanced dataset of misogynous and non-misogynous tweets. The raw training set (6,000 tweets) is derived from the data collected for the 2018 edition of the AMI shared task, where the misogynistic posts have been enriched by labelling aggressive expressions according to the definition provided in Section 2. The raw test dataset (approximately 1,000 tweets) has been collected from Twitter using a similar approach to the 2018 edition of the shared task. This is intentionally done to evaluate the generalization abilities of the systems on test data collected in a different time period and therefore characterized by higher language variability with respect to the training data. Examples of tweets belonging to the raw dataset are shown in Table 1.

Table 1: Examples of misogynous and non-misogynous tweets

Misogynous

Aggressiveness

Text

1

1

Stupida tr**a! Meriti di morire

1

0

Le ragazze sono tutte pu***ne

0

0

@chiellini sei una cagna

8The training raw data for this dataset are provided as TSV files (tab-separated files) and report the following fields, where:

  • id denotes a unique identifier of the tweet.

  • text represents the tweet text.

  • misogynous defines whether a tweet is misogynous or not misogynous; it takes values:
    – 0 if the tweet is not misogynous;
    – 1 if the tweet is misogynous.

  • aggressiveness denotes whether a misogynous tweet is aggressive or not; it takes value as:
    – 0 denotes a non-aggressive tweet (not misogynous tweets are labelled as 0 by default);
    – 1 if the tweet is aggressive.

9The raw testing data are provided as TSV files reporting only id and text.

3.2 Synthetic dataset

10The synthetic test dataset for measuring the presence of unintended bias has been created following the procedure adopted in (Dixon et al. 2018; Nozza, Volpetti, and Fersini 2019): a list of identity terms has been constructed by taking into consideration some concepts related to the term “donna" (e.g. “moglie", “fidanzata"). Given the identity terms, several templates have been created including positive/negative verbs and adjectives (e.g. negative: hate, inferior; positive: love, awesome) both for conveying a misogynistic message or a non-misogynistic one. Some examples of such templates, used to create the synthetic dataset, are reported in Table 2.

Table 2. Template examples

Template Examples

Misogynous

<identity_term>dovrebbero essere protette

0

<identity_term>dovrebbero essere uccise

1

apprezziamo <identity_term>

0

picchiamo <identity_term>

1

splendida <identity_term>

0

sporca <identity_term>

1

11The synthetic dataset, created for measuring the presence of unintended bias, contains template-generated text labelled according to:

  • Misogyny: Misogyny (1) vs. Not Misogyny (0)

12The training data for the raw dataset are provided as TSV files (tab-separated files) and report the following fields:

  • id denotes a unique identifier of the template-generated text.

  • text represents the template-generated text.

  • misogynous defines if the template-generated text is misogynous or non-misogynous; it takes values as 1 if the tweet is misogynous, 0 if the tweet is non-misogynous.

13The synthetic testing data are provided as TSV files (tab-separated files) reporting only id and text.

14The statistics about the raw and synthetic datasets, both for the training and testing sets, are reported in Table 3.

Table 3. Distribution of labels on the Training and Test datasets

Training

Testing

Raw

Synthetic

Raw

Synthetic

Misogynous

2337

1007

500

954

Non-misogynous

2663

1007

500

954

Aggressive

1783

-

176

-

Non-aggressive

3217

-

824

-

4. Evaluation Measures and Baseline

15Considering the distribution of labels of the dataset, we have chosen different evaluation metrics. In particular, we distinguished as follows:

16Subtask A. Each class to be predicted (i.e. “Misogyny" and “Aggressiveness") has been evaluated independently on the other using a Macro F1-score. The final ranking of the systems participating in Subtask A was based on the Average Macro F1-score (F1), computed as follows:

17Subtask B. The ranking for Subtask B is computed by the weighted combination of AUC estimated on the test raw dataset AUCraw and three per-term AUC-based bias scores computed on the synthetic dataset (AUCSubgroup, AUCBPSN, AUCBNSP). Let s be an identity-term (e.g. “girlfriend" and “wife") and N be the total number of identity-terms, the score of each run is estimated according to the following metric:

18Unintended bias can be uncovered by looking at differences in the score distributions between data mentioning a specific identity-term (subgroup distribution) and the rest (background distribution). The three per-term AUC-based bias scores are related to specific subgroups as follows:

  • AUCSubgroup(s): calculates AUC only on the data within the subgroup related to a given identity term. This represents model understanding and separability within the subgroup itself. A low value in this metric means the model does a poor job of distinguishing between misogynous and non-misogynous comments that mention the identity.

  • AUCBPSN(s): Background Positive Subgroup Negative (BPSN) calculates AUC on the misogynous examples from the background and the non-misogynous examples from the subgroup. A low value in this metric means that the model confuses non-misogynous examples that mention the identity-term with misogynous examples that do not, likely meaning that the model predicts higher misogynous scores than it should for non-misogynous examples mentioning the identity-term.

  • AUCBNSP(s): Background Negative Subgroup Positive (BNSP) calculates AUC on the non-misogynous examples from the background and the misogynous examples from the subgroup. A low value here means that the model confuses misogynous examples that mention the identity with non-misogynous examples that do not, likely meaning that the model predicts lower misogynous scores than it should for misogynous examples mentioning the identity.

19In order to compare the submitted runs with a baseline model, we provided a benchmark (AMI-BASELINE) based on Support Vector Machine trained on a unigram representation of tweets with Tf-IDF weighing schema. In particular, we created one training set for each field to be predicted, i.e. “misogynous”, “aggressiveness”, where each tweet has been represented as a bag-of-words (composed of 1000 terms) coupled with the corresponding label. Once the representations have been obtained, Support Vector Machines with linear kernel have been trained and provided as AMI-BASELINE.

5. Participants and Results

20A total of 8 teams from 6 different countries participated in at least one of the two subtasks of AMI. Two teams participated with the same approach also in the HaSpeeDe shared task (Sanguinetti et al. 2020), addressing misogyny identification with generic models for detecting hate speech. Each team had the chance to submit up to three runs that could be constrained (c), where only the provided training data and lexicons were admitted, and unconstrained (u), where additional data for training were allowed. Table 4 reports an overview of the teams illustrating their affiliation, their country, the number and type (c for constrained, u for unconstrained) of submissions, and the subtasks they addressed.

5.1 Subtask A: Misogyny & Aggressive Behaviour Identification

21Table 5 reports the results for the Misogyny & Aggressive Behaviour Identification task, which received 20 submissions submitted by 8 teams. The highest result has been achieved by jigsaw at 0.7406 in an unconstrained setting and by fabsam at 0.7342 in a constrained run. While the best results obtained as unconstrained is based on ensembles of fine-tuned custom BERT models, the one achieved by the best constrained system is grounded on a convolutional neural network that exploits pre-trained word embeddings.

22By analysing the detailed results, it emerged that while the identification of misogynous text can be considered a quite simple problem, the recognition of aggressiveness needs to be properly addressed. In fact, the score reported in Table 5 are strongly affected by the prediction capabilities mostly related to the aggressive posts. This is likely due to the subjective perception of aggressiveness captured by the variance of the data available in the ground truth.

Table 5. Results of Subtask A. Constrained runs are marked as “c", while the unconstrained ones with “u". An amended run, marked with **, has been submitted after the deadline

Rank

Run Type

Score

Team

**

c

0.744

UniBO **

1

u

0.741

jigsaw

2

u

0.738

jigsaw

3

c

0.734

fabsam

4

u

0.731

YNU_OXZ

5

c

0.731

fabsam

6

c

0.717

NoPlaceForHateSpeech

7

u

0.701

YNU_OXZ

8

c

0.695

fabsam

9

c

0.693

NoPlaceForHateSpeech

10

c

0.687

AMI_the_winner

11

u

0.684

MDD

12

c

0.683

PoliTeam

13

c

0.682

MDD

14

c

0.681

PoliTeam

15

u

0.668

MDD

16

c

0.665

AMI_the_winner

17

c

0.665

AMI_BASELINE

18

c

0.647

PoliTeam

19

c

0.634

UniBO

20

c

0.626

AMI_the_winner

21

c

0.490

NoPlaceForHateSpeech

23After the deadline the team UniBO submitted an amended run (**), that has not been ranked in the official results of the AMI shared task. However, we believe interesting to mention their achievement showing an Average Macro F1-score equal to 0.744.

5.2 Subtask B: Unbiased Misogyny Identification

24Table 5 reports the results for the Unbiased Misogyny Identification task, which received 11 submissions by 4 teams, among which 4 unconstrained and 7 constrained. The highest Average Macro F1 score has been achieved by jigsaw at 0.8825 with an unconstrained run and by PoliTeam at 0.8180 with a constrained submission.

Table 6. Results of Subtask B. Constrained runs are marked as “c", while the unconstrained ones with “u"

Rank

Run Type

Score

Team

1

u

0.882

jigsaw

2

c

0.818

PoliTeam

3

c

0.814

PoliTeam

4

c

0.705

fabsam

5

c

0.702

fabsam

6

c

0.694

PoliTeam

7

c

0.691

fabsam

8

u

0.649

jigsaw

9

c

0.613

MDD

10

c

0.602

AMI_BASELINE

11

u

0.601

MDD

12

u

0.601

MDD

25Similarly to the previous task, most of the systems have shown better performance compared to the AMI-BASELINE. By analizing the runs, we can highlight that the two best results achieved on Subtask B have been obtained by the unconstrained run submitted by jigsaw, where a simple debiasing technique based on data augumentation have been adopted, and by the constrained run provided by Politeam, where the problem of biased prediction has been partially mitigated by introducing misogynous lexicon.

6. Discussion

26The submitted systems can be compared by taking into consideration the kind of input feature that they have considered for representing tweets and the machine learning model that has been used as classification model.

Textual Feature Representation

27The systems submitted by the challenge participants’ consider various techniques for representing the tweet contents. Most of the teams experimented a high-level representation of the text based deep learning solutions. While few teams like fabsam and MDD adopted a text representation based on traditional word embeddings such as Word2Vec (Mikolov et al. 2013), Glove (Pennington, Socher, and Manning 2014) and FastText (Bojanowski et al. 2017), most of the systems. i.e NoPlaceForHateSpeech,jigsaw, PoliTeam, YNU_OXZ and UniBO, exploited richer sentence embeddings such as BERT (Devlin et al. 2019) or XLM-RoBert (Ruder et al., 2019). For enriching the space for then training the subsequent models to recognize misogyny and aggressiveness, PoliTeam experimented the use of additional lexical resources such as misogynous lexicon and sentiment Lexicon.

Machine Learning Models

28Concerning the machine learning models, we can distinguish between approaches trained from scratch and those ones based on fine-tuning of existing pre-trained models. We report in the following the strategy adopted by the systems that participated in the AMI shared task, according to the type of machine learning model that has been adopted:

  • Shallow models have been experimented by MDD, where logistic regressions have been trained according to different hand-crafted features;

  • Convolutional Neural Networks have been exploited by NoPlaceForHateSpeech by using two distinct models for misogyny detection and aggressiveness identification, by fabsal investigating the optimal hyperparameters of the model, and by YNU_OXZ where on top of the CNN architecture a Capsule Network (Sabour, Frosst, and Hinton 2017) has been introduced for taking advantage of spatial patterns available in short texts;

  • Fine-Tuning of pre-trained models has been exploited by jigsaw by adapting BERT to the challenge domain and using a transfer multilingual strategy and ensemble learning, by UniBO that accommodated the BERT model using a multi-label output neuron, and by PoliTeam where the prediction of the fine-tuned sentence-BERT is coupled with prediction based on lexicons.

29For what concerns the achieved results on the two subtasks, few considerations can be drawn considering both the errors done by the systems and the mitigation strategies adopted for reducing the bias.

Error Analysis

30When testing the developed systems on raw test data, the majority of the performed errors can be summarized by the following patterns:

  • Under-representation of subjective expressions: those posts written by introducing erroneous lower case and missing spaces between adjoining words lead the models based on raw text to make errors on test predictions. An example of such common errors is the one reported in the following tweet:
    “Odio Sakura per il semplice motivo che qualunque cosa faccia o dica Naruto lei lo prende a schiaffi o a pugniHA CHIESTO COME STA SAI DIOSANTO BRUTTA STRONZA MA CONTRALLI MADONNA SPERO CHE TI UCCIDANOscusami Sarada"

  • Undefined subject, but presence of aggressive terms: for those tweets where the target is not clearly mentioned, but several aggressive terms are present, the models tend to be biased and to predict the post as misogynous and aggressive erroneously. An example of this type of misclassified posts is reported here:
    “Vero...ma c’e chi ti cerca, che ti vuole, più di ogni cosa al mondo......ma non sa se viene capito..... potrebbe esser mal interpretato e di conseguenza all’abbraccio esser denunciato per molestie sessuali e/o stupro"

  • Short hate speech sentences referred to others than women: when the target is mentioned by using an actual account, but it is referred to men, and there are no additional indications about the gender of the target, most of the models tend to misclassify the tweet. In the following example, the target is a male football player:
    “@bonucci_leo19 Cagati in mano e prenditi a schiaffi. Sti post te li infili nel c*lo!"
    Concerning the errors on the synthetic test set used for estimating the bias of the models, two main errors carried out by the majority of the systems can be identified:

  • presence of unusual target: in most of the submissions, sentences containing offensive expressions towards specific uncommon targets are misclassified. For instance, around 39% of the predictions related to the target nonna (i.e., grandmother) are wrong. An example of the most misclassified target is reported in the following example:
    “nonne belle"

  • Presence of unusual verbs: analogously to what has been observed for the target, sentences containing rare aggressive verbs tend to be misclassified. For instance, around 48% of the instances related to the verbs malmenare and seviziare (i.e., beat up and torture) are wrongly classified. An example of a mistaken sentence are reported here:
    “femmina dovrebbe essere seviziata" (wrongly classified as non-misogynous)

Bias Mitigation strategies.

31Concerning the Subtask B, only one team (jigsaw) addressed explicitly the problem related to the unintended bias. The authors used sentences sampled from the Italian Wikipedia articles containing some of the identity terms provided with the test set. These sentences, labeled as both non-misogynous and non-aggressive, have been used to further fine-tune the model and reduce the bias given by the data. The results achieved by the jigsaw team highlight that a debiasing method could obtain fair predictions even using pre-trained models.

7. Conclusions and Future Work

32This paper presents the AMI shared task, focused not only on identifying misogynous and aggressive expressions but also on ensuring fair predictions. By analysing the runs submitted by the participants, we can conclude that while the problem of misogyny identification has reached satisfactory results, the recognition of aggressiveness is still in its infancy. Concerning the capabilities of the systems with respect to the unintended bias problem, we can highlight that a domain-dependent mitigation strategy is a necessary step towards fair models.

Literaturverzeichnis

Resham Ahluwalia, Himani Soni, Edward Callow, Anderson Nascimento, and Martine De Cock. 2018. “Detecting Hate Speech Against Women in English Tweets.” In Proceedings of Sixth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2018). Turin, Italy: CEUR.org.

Maria Anzovino, Elisabetta Fersini, and Paolo Rosso. 2018. “Automatic Identification and Classification of Misogynistic Language on Twitter.” In Proceedings of 23rd International Conference on Applications of Natural Language to Information Systems (NLDB), 57–64. Springer.

Giuseppe Attanasio and Eliana Pastor. 2020. PoliTeam@ AMI: Improving Sentence Embedding Similarity with Misogyny Lexicons for Automatic Misogyny Identification in Italian Tweets. In Proceedings of Seventh Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Work-shop (EVALITA 2020), Bologna, Italy. CEUR.org.

Amir Bakarov. 2018. “Vector Space Models for Automatic Misogyny Identification.” In Proceedings of Sixth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2018). Turin, Italy: CEUR.org.

Angelo Basile and Chiara Rubagotti. 2018. “Automatic Identification of Misogyny in English and Italian Tweets at EVALITA 2018 with a Multilingual Hate Lexicon.” In Proceedings of Sixth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2018). Turin, Italy: CEUR.org.

Valerio Basile, Cristina Bosco, Elisabetta Fersini, Nozza Debora, Viviana Patti, Francisco Manuel Rangel Pardo, Paolo Rosso, Manuela Sanguinetti, and others. 2019. “Semeval-2019 Task 5: Multilingual Detection of Hate Speech Against Immigrants and Women in Twitter.” In Proceedings of 13th International Workshop on Semantic Evaluation, 54–63. Association for Computational Linguistics.

Valerio Basile, Danilo Croce, Maria Di Maro, and Lucia C. Passaro. 2020. “EVALITA 2020: Overview of the 7th Evaluation Campaign of Natural Language Processing and Speech Tools for Italian.” In Proceedings of Seventh Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (Evalita 2020), edited by Valerio Basile, Danilo Croce, Maria Di Maro, and Lucia C. Passaro. Online: CEUR.org.

Piotr Bojanowski, Edouard Grave, Armand Joulin, and Tomas Mikolov. 2017. “Enriching Word Vectors with Subword Information.” Transactions of the Association for Computational Linguistics 5: 135–46.

Davide Buscaldi. 2018. “Tweetaneuse AMI EVALITA2018: Character-based Models for the Automatic Misogyny Identification Task.” In Proceedings of Sixth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2018). Turin, Italy: CEUR.org.

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. “BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding.” In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), 4171–86. Association for Computational Linguistics. https://doi.org/10.18653/v1/n19-1423.

Lucas Dixon, John Li, Jeffrey Sorensen, Nithum Thain, and Lucy Vasserman. 2018. “Measuring and Mitigating Unintended Bias in Text Classification.” In Proceedings of the 2018 Aaai/Acm Conference on Ai, Ethics, and Society, 67–73.

Samer El Abassi and Sergiu Nisioi. 2020. MDD@AMI: Vanilla Classifiers for Misogyny Identification. In Proceedings of Sixth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2020), Bologna, Italy. CEUR.org.

Samuel Fabrizi. 2020. fabsam @ AMI: a Convolutional Neural Network approach. In Proceedings of Seventh Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Work-shop (EVALITA 2020), Bologna, Italy. CEUR.org.

Elisabetta Fersini, Debora Nozza, and Paolo Rosso. 2018. “Overview of the Evalita 2018 Task on Automatic Misogyny Identification (AMI).” In Proceedings of the Sixth evaluation campaign of Natural Language Processing and Speech tools for Italian (EVALITA 2018), edited by Tommaso Caselli, Nicole Novielli, Viviana Patti, and Paolo Rosso. Turin, Italy: CEUR.org.

Elisabetta Fersini, Paolo Rosso, and Maria Anzovino. 2018. “Overview of the Task on Automatic Misogyny Identification at IberEval 2018.” In IberEval@ Sepln, 214–28.

Simona Frenda, Bilal Ghanem, Estefanı́a Guzmán-Falcón, Manuel Montes-y-Gómez, and Luis Villaseñor-Pineda. 2018. “Automatic Lexicons Expansion for Multilingual Misogyny Detection.” In Proceedings of Sixth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2018). Turin, Italy: CEUR.org.

Sarah Hewitt, Thanassis Tiropanis, and Christian Bokhove. 2016. The Problem of identifying Misog-ynist Language on Twitter (and other online social spaces). In Proceedings of the 8th ACM Conference on Web Science, pages 333–335. ACM.

Alyssa Lees, Jeffrey Sorensen, and Ian Kivlichan. 2020. Jigsaw @ AMI and HaSpeeDe2: Fine-Tuning a Pre-Trained Comment-Domain BERT Model. In Proceedings of Seventh Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2020), Bologna, Italy. CEUR.org.

Marco Lepri, Giuseppe Grieco, and Mattia Sangermano. University of Pisa, Italy.

Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg S Corrado, and Jeff Dean. 2013. “Distributed Representations of Words and Phrases and Their Compositionality.” In Advances in Neural Information Processing Systems, 3111–9.

Arianna Muti and Alberto Barr`on-Cedeño. 2020. UniBO@AMI: A Multi-Class Approach to Misogyny and Aggressiveness Identification on Twitter Posts Using AlBERTo. In Proceedings of Seventh Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Work-shop (EVALITA 2020), Bologna, Italy. CEUR.org.

Debora Nozza, Claudia Volpetti, and Elisabetta Fersini. 2019. “Unintended Bias in Misogyny Detection.” In IEEE/Wic/Acm International Conference on Web Intelligence, 149–55.

Xiaozhi Ou and Hongling Li. 2020. YNU OXZ @ HaSpeeDe 2 and AMI : XLM-RoBERTa with Ordered Neurons LSTM for classification task at EVALITA 2020. In Proceedings of Sixth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2020), Bologna, Italy. CEUR.org.

Endang Wahyu Pamungkas, Alessandra Teresa Cignarella, Valerio Basile, and Viviana Patti. 2018. “Automatic Identification of Misogyny in English and Italian Tweets at EVALITA 2018 with a Multilingual Hate Lexicon.” In Proceedings of Sixth Evaluation Campaign of Natural Language Processing and Speech Tools for Italian. Final Workshop (EVALITA 2018). Turin, Italy: CEUR.org.

Jeffrey Pennington, Richard Socher, and Christopher D. Manning. 2014. “GloVe: Global Vectors for Word Representation.” In Empirical Methods in Natural Language Processing, 1532–43. http://www.aclweb.org/anthology/D14-1162.

Bailey Poland. 2016. Haters: Harassment, Abuse, and Violence Online. Potomac Books, Incorporated.

Sebastian Ruder, Anders Søgaard, and Ivan Vulić 2019. Unsupervised cross-lingual representation learning. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: Tutorial Abstracts, pages 31–38. Association for Computational Linguistics.

Sara Sabour, Nicholas Frosst, and Geoffrey E Hinton. 2017. “Dynamic Routing Between Capsules.” In Advances in Neural Information Processing Systems, 3856–66.

Manuela Sanguinetti, Gloria Comandini, Elisa Di Nuovo, Simona Frenda, Marco Stranisci, Cristina Bosco, Tommaso Caselli, Viviana Patti, and Irene Russo. 2020. “HaSpeeDe 2@EVALITA2020: Overview of the EVALITA 2020 Hate Speech Detection Task.” In Proceedings of the 7th evaluation campaign of Natural Language Processing and Speech tools for Italian (EVALITA 2020), edited by Valerio Basile, Danilo Croce, Maria Di Maro, and Lucia C. Passaro. Online: CEUR.org.

Autoren

CC-BY-NC-ND-4.0

Nur der Text ist unter der Lizenz CC BY-NC-ND 4.0 nutzbar. Alle anderen Elemente (Abbildungen, importierte Anhänge) sind „Alle Rechte vorbehalten“, sofern nicht anders angegeben.

Lesen

Open access

Suche in OpenEdition Search

Sie werden weitergeleitet zur OpenEdition Search