Natural Language Processing (NLP) models are found to exhibit discriminatory stereotypes across many social constructs, e.g. gender and race. In comparison to the progress made in reducing bias from static word embeddings, fairness in sentence-level text encoders received little consideration despite their wider applicability in contemporary NLP tasks. In this paper, we propose a debiasing method for pre-trained text encoders that both reduces social stereotypes, and inflicts next to no semantic damage. Unlike previous studies that directly manipulate the embeddings, we suggest to dive deeper into the operation of these encoders, and pay more attention to the way they pay attention to different social groups. We find that stereotypes are also encoded in the attention layer. Then, we work on model debiasing by redistributing the attention scores of a text encoder such that it forgets any preference to historically advantaged groups, and attends to all social classes with the same intensity. Our experiments confirm that reducing bias from attention effectively mitigates it from the model’s text representations.

Tagging in the context of online resources is a fundamental addition to search systems. Tags assist with the indexing, management, and retrieval of online products and services to answer complex user queries. Traditional methods of matching user queries with tags either rely on cosine similarity, or employ semantic similarity models that fail to recognize conceptual connections between tags, e.g. ambiance and music. In this work, we focus on subjective tags which characterize subjective aspects of a product or service. We propose conceptual similarity to leverage conceptual awareness when assessing similarity between tags. We also provide a simple cost-effective pipeline to automatically generate data in order to train the conceptual similarity model. We show that our pipeline generates high-quality datasets, and evaluate the similarity model both systematically and on a downstream application. Experiments show that conceptual similarity outperforms existing work when using subjective tags.

Developing bots demands highquality training samples, typically in the form of user utterances and their associated intents. Given the fuzzy nature of human language, such datasets ideally must cover all possible utterances of each single intent. Crowdsourcing has widely been used to collect such inclusive datasets by paraphrasing an initial utterance. However, the quality of this approach often suffers from various issues, particularly language errors produced by unqualified crowd workers. More so, since workers are tasked to write open-ended text, it is very challenging to automatically asses the quality of paraphrased utterances. In this paper, we investigate common crowdsourced paraphrasing issues, and propose an annotated dataset called Para-Quality, for detecting the quality issues. We also investigate existing tools and services to provide baselines for detecting each category of issues. In all, this work presents a data-driven view of incorrect paraphrases during the bot development process, and we pave the way towards automatic detection of unqualified paraphrases.