1 |
Disambiguation of morpho-syntactic features of African American English -- the case of habitual be ...
|
|
|
|
Abstract:
Recent research has highlighted that natural language processing (NLP) systems exhibit a bias against African American speakers. The bias errors are often caused by poor representation of linguistic features unique to African American English (AAE), due to the relatively low probability of occurrence of many such features in training data. We present a workflow to overcome such bias in the case of habitual "be". Habitual "be" is isomorphic, and therefore ambiguous, with other forms of "be" found in both AAE and other varieties of English. This creates a clear challenge for bias in NLP technologies. To overcome the scarcity, we employ a combination of rule-based filters and data augmentation that generate a corpus balanced between habitual and non-habitual instances. With this balanced corpus, we train unbiased machine learning classifiers, as demonstrated on a corpus of AAE transcribed texts, achieving .65 F$_1$ score disambiguating habitual "be". ...
|
|
Keyword:
Computation and Language cs.CL; FOS Computer and information sciences
|
|
URL: https://dx.doi.org/10.48550/arxiv.2204.12421 https://arxiv.org/abs/2204.12421
|
|
BASE
|
|
Hide details
|
|
3 |
Supplementary material from " DCDC2 READ1 regulatory element: how temporal processing differences may shape language" ...
|
|
|
|
BASE
|
|
Show details
|
|
4 |
Supplementary material from " DCDC2 READ1 regulatory element: how temporal processing differences may shape language" ...
|
|
|
|
BASE
|
|
Show details
|
|
5 |
Data from: DCDC2 READ1 regulatory element: how temporal processing differences may shape language ...
|
|
|
|
BASE
|
|
Show details
|
|
6 |
Supplementary Materials for DCDC2 READ1 regulatory element: how temporal processing differences may shape language ...
|
|
|
|
BASE
|
|
Show details
|
|
7 |
Supplementary Materials for DCDC2 READ1 regulatory element: how temporal processing differences may shape language ...
|
|
|
|
BASE
|
|
Show details
|
|
8 |
DCDC2 READ1 regulatory element: how temporal processing differences may shape language
|
|
|
|
In: Proc Biol Sci (2020)
|
|
BASE
|
|
Show details
|
|
9 |
Does Working-Memory Training Given to Reception-Class Children Improve the Speech of Children at Risk of Fluency Difficulty?
|
|
|
|
In: Front Psychol (2020)
|
|
BASE
|
|
Show details
|
|
10 |
Statistical and acoustic effects on the perception of stop consonants in Kaqchikel (Mayan)
|
|
|
|
In: Laboratory Phonology: Journal of the Association for Laboratory Phonology; Vol 9, No 1 (2018); 9 ; 1868-6354 (2018)
|
|
BASE
|
|
Show details
|
|
|
|