In recent years, word embeddings have been widely used to measure biases in texts. Even if they have proven to be effective in detecting a wide variety of biases, metrics based on word embeddings lack transparency and interpretability. We analyze an alternative PMI-based metric to quantify biases in texts. It can be expressed as a function of conditional probabilities, which provides a simple interpretation in terms of word co-occurrences. We also prove that it can be approximated by an odds ratio, which allows estimating confidence intervals and statistical significance of textual biases. This approach produces similar results to metrics based on word embeddings when capturing gender gaps of the real world embedded in large corpora.

Recent research has shown that static word embeddings can encode words’ frequencies. However, little has been studied about this behavior. In the present work, we study how frequency and semantic similarity relate to one another in static word embeddings, and we assess the impact of this relationship on embedding-based bias metrics. We find that Skip-gram, GloVe and FastText embeddings tend to produce higher similarity between high-frequency words than between other frequency combinations. We show that the association between frequency and similarity also appears when words are randomly shuffled, and holds for different hyperparameter settings. This proves that the patterns we find are neither due to real semantic associations nor to specific parameters choices, and are an artifact produced by the word embeddings. To illustrate how frequencies can affect the measurement of biases related to gender, ethnicity, and affluence, we carry out a controlled experiment that shows that biases can even change sign or reverse their order when word frequencies change.

Numerous works use word embedding-based metrics to quantify societal biases and stereotypes in texts. Recent studies have found that word embeddings can capture semantic similarity but may be affected by word frequency. In this work we study the effect of frequency when measuring female vs. male gender bias with word embedding-based bias quantification methods. We find that Skip-gram with negative sampling and GloVe tend to detect male bias in high frequency words, while GloVe tends to return female bias in low frequency words. We show these behaviors still exist when words are randomly shuffled. This proves that the frequency-based effect observed in unshuffled corpora stems from properties of the metric rather than from word associations. The effect is spurious and problematic since bias metrics should depend exclusively on word co-occurrences and not individual word frequencies. Finally, we compare these results with the ones obtained with an alternative metric based on Pointwise Mutual Information. We find that this metric does not show a clear dependence on frequency, even though it is slightly skewed towards male bias across all frequencies.

Mental health forums are online spaces where people can share their experiences anonymously and get peer support. These forums, require the supervision of moderators to provide support in delicate cases, such as posts expressing suicide ideation. The large increase in the number of forum users makes the task of the moderators unmanageable without the help of automatic triage systems. In the present paper, we present a Machine Learning approach for the triage of posts. Most approaches in the literature focus on the content of the posts, but only a few authors take advantage of features extracted from the context in which they appear. Our approach consists of the development and implementation of a large variety of new features from both, the content and the context of posts, such as previous messages, interaction with other users and author’s history. Our method has competed in the CLPsych 2017 Shared Task, obtaining the first place for several of the subtasks. Moreover, we also found that models that take advantage of post context improve significantly its performance in the detection of flagged posts (posts that require moderators attention), as well as those that focus on post content outperforms in the detection of most urgent events.

Despite the popularity of word embeddings, the precise way by which they acquire semantic relations between words remain unclear. In the present article, we investigate whether LSA and word2vec capacity to identify relevant semantic relations increases with corpus size. One intuitive hypothesis is that the capacity to identify relevant associations should increase as the amount of data increases. However, if corpus size grows in topics which are not specific to the domain of interest, signal to noise ratio may weaken. Here we investigate the effect of corpus specificity and size in word-embeddings, and for this, we study two ways for progressive elimination of documents: the elimination of random documents vs. the elimination of documents unrelated to a specific task. We show that word2vec can take advantage of all the documents, obtaining its best performance when it is trained with the whole corpus. On the contrary, the specialization (removal of out-of-domain documents) of the training corpus, accompanied by a decrease of dimensionality, can increase LSA word-representation quality while speeding up the processing time. From a cognitive-modeling point of view, we point out that LSA’s word-knowledge acquisitions may not be efficiently exploiting higher-order co-occurrences and global relations, whereas word2vec does.