| 1. |
- Adelani, David Ifeoluwa, et al.
(författare)
-
MasakhaNER: Named Entity Recognition for African Languages
- 2021
-
Ingår i: Transactions of the Association for Computational Linguistics. - : MIT Press. - 2307-387X. ; 9, s. 1116-1131
-
Tidskriftsartikel (refereegranskat)abstract
- We take a step towards addressing the under-representation of the African continent in NLP research by bringing together different stakeholders to create the first large, publicly available, high-quality dataset for named entity recognition (NER) in ten African languages. We detail the characteristics of these languages to help researchers and practitioners better understand the challenges they pose for NER tasks. We analyze our datasets and conduct an extensive empirical evaluation of state-of-the-art methods across both supervised and transfer learning settings. Finally, we release the data, code, and models to inspire future research on African NLP.
|
|
| 2. |
- Ampoorter, Evy, et al.
(författare)
-
Tree diversity is key for promoting the diversity and abundance of forest-associated taxa in Europe
- 2020
-
Ingår i: Oikos. - : Wiley. - 0030-1299 .- 1600-0706. ; 129:2, s. 133-146
-
Tidskriftsartikel (refereegranskat)abstract
- Plant diversity is an important driver of diversity at other trophic levels, suggesting that cascading extinctions could reduce overall biodiversity. Most evidence for positive effects of plant diversity comes from grasslands. Despite the fact that forests are hotspots of biodiversity, the importance of tree diversity, in particular its relative importance compared to other management related factors, in affecting forest-associated taxa is not well known. To address this, we used data from 183 plots, located in different forest types, from Mediterranean to Boreal, and established along a climatic gradient across six European countries (FunDivEUROPE project). We tested the influence of tree diversity, tree functional composition (i.e. functional trait values), forest structure, climate and soil on the diversity and abundance/activity of nine taxa (bats, birds, spiders, microorganisms, earthworms, ungulates, foliar fungal pathogens, defoliating insects and understorey plants) and on their overall diversity and abundance/activity (multidiversity, multiabundance/activity). Tree diversity was a key driver of taxon-level and overall forest-associated biodiversity, along with tree functional composition, forest structure, climate and soil. Both tree species richness and functional diversity (variation in functional trait values) were important. The effects of tree diversity on the abundance/activity of forest-associated taxa were less consistent. Nonetheless, spiders, ungulates and foliar fungal pathogens were all more abundant/active in diverse forests. Tree functional composition and structure were also important drivers of abundance/activity: conifer stands had lower overall multidiversity (although the effect was driven by defoliating insects), while stands with potentially tall trees had lower overall multiabundance/activity. We found more synergies than tradeoffs between diversity and abundance/activity of different taxa, suggesting that forest management can promote high diversity across taxa. Our results clearly show the high value of mixed forest stands for multiple forest-associated taxa and indicate that multiple dimensions of tree diversity (taxonomic and functional) are important.
|
|
| 3. |
- Baeten, Lander, et al.
(författare)
-
Identifying the tree species compositions that maximize ecosystem functioning in European forests
- 2019
-
Ingår i: Journal of Applied Ecology. - : Wiley. - 0021-8901 .- 1365-2664. ; 56:3, s. 733-744
-
Tidskriftsartikel (refereegranskat)abstract
- 1. Forest ecosystem functioning generally benefits from higher tree species richness, but variation within richness levels is typically large. This is mostly due to the contrasting performances of communities with different compositions. Evidence-based understanding of composition effects on forest productivity, as well as on multiple other functions will enable forest managers to focus on the selection of species that maximize functioning, rather than on diversity per se.2. We used a dataset of 30 ecosystem functions measured in stands with different species richness and composition in six European forest types. First, we quantified whether the compositions that maximize annual above-ground wood production (productivity) generally also fulfil the multiple other ecosystem functions (multifunctionality). Then, we quantified the species identity effects and strength of interspecific interactions to identify the "best" and "worst" species composition for multifunctionality. Finally, we evaluated the real-world frequency of occurrence of best and worst mixtures, using harmonized data from multiple national forest inventories.3. The most productive tree species combinations also tended to express relatively high multifunctionality, although we found a relatively wide range of compositions with high- or low-average multifunctionality for the same level of productivity. Monocultures were distributed among the highest as well as the lowest performing compositions. The variation in functioning between compositions was generally driven by differences in the performance of the component species and, to a lesser extent, by particular interspecific interactions. Finally, we found that the most frequent species compositions in inventory data were monospecific stands and that the most common compositions showed below-average multifunctionality and productivity.4. Synthesis and applications. Species identity and composition effects are essential to the development of high-performing production systems, for instance in forestry and agriculture. They therefore deserve great attention in the analysis and design of functional biodiversity studies if the aim is to inform ecosystem management. A management focus on tree productivity does not necessarily trade-off against other ecosystem functions; high productivity and multifunctionality can be combined with an informed selection of tree species and species combinations.
|
|
| 4. |
- Beloucif, Meriem, et al.
(författare)
-
Elvis vs. M. Jackson : Who has More Albums? Classification and Identification of Elements in Comparative Questions
- 2022
-
Ingår i: LREC 2022. - : European Language Resources Association. - 9791095546726 ; , s. 3771-3779
-
Konferensbidrag (refereegranskat)abstract
- Comparative Question Answering (cQA) is the task of providing concrete and accurate responses to queries such as: "Is Lyft cheaper than a regular taxi?" or "What makes a mortgage different from a regular loan?". In this paper, we propose two new open-domain real-world datasets for identifying and labeling comparative questions. While the first dataset contains instances of English questions labeled as comparative vs. non-comparative, the second dataset provides additional labels including the objects and the aspects of comparison. We conduct several experiments that evaluate the soundness of our datasets. The evaluation of our datasets using various classifiers show promising results that reach close-to-human results on a binary classification task with a neural model using ALBERT embeddings. When approaching the unsupervised sequence labeling task, some headroom remains.
|
|
| 5. |
- Blalock, Zachary N., et al.
(författare)
-
Circulating cell-free mitochondrial DNA levels and glucocorticoid sensitivity in a cohort of male veterans with and without combat-related PTSD
- 2024
-
Ingår i: Translational Psychiatry. - 2158-3188. ; 14:1
-
Tidskriftsartikel (refereegranskat)abstract
- Circulating cell-free mitochondrial DNA (ccf-mtDNA) is a biomarker of cellular injury or cellular stress and is a potential novel biomarker of psychological stress and of various brain, somatic, and psychiatric disorders. No studies have yet analyzed ccf-mtDNA levels in post-traumatic stress disorder (PTSD), despite evidence of mitochondrial dysfunction in this condition. In the current study, we compared plasma ccf-mtDNA levels in combat trauma-exposed male veterans with PTSD (n = 111) with those who did not develop PTSD (n = 121) and also investigated the relationship between ccf mt-DNA levels and glucocorticoid sensitivity. In unadjusted analyses, ccf-mtDNA levels did not differ significantly between the PTSD and non-PTSD groups (t = 1.312, p = 0.191, Cohen’s d = 0.172). In a sensitivity analysis excluding participants with diabetes and those using antidepressant medication and controlling for age, the PTSD group had lower ccf-mtDNA levels than did the non-PTSD group (F(1, 179) = 5.971, p = 0.016, partial η 2 = 0.033). Across the entire sample, ccf-mtDNA levels were negatively correlated with post-dexamethasone adrenocorticotropic hormone (ACTH) decline (r = −0.171, p = 0.020) and cortisol decline (r = −0.149, p = 0.034) (viz., greater ACTH and cortisol suppression was associated with lower ccf-mtDNA levels) both with and without controlling for age, antidepressant status and diabetes status. Ccf-mtDNA levels were also significantly positively associated with IC50-DEX (the concentration of dexamethasone at which 50% of lysozyme activity is inhibited), a measure of lymphocyte glucocorticoid sensitivity, after controlling for age, antidepressant status, and diabetes status (β = 0.142, p = 0.038), suggesting that increased lymphocyte glucocorticoid sensitivity is associated with lower ccf-mtDNA levels. Although no overall group differences were found in unadjusted analyses, excluding subjects with diabetes and those taking antidepressants, which may affect ccf-mtDNA levels, as well as controlling for age, revealed decreased ccf-mtDNA levels in PTSD. In both adjusted and unadjusted analyses, low ccf-mtDNA levels were associated with relatively increased glucocorticoid sensitivity, often reported in PTSD, suggesting a link between mitochondrial and glucocorticoid-related abnormalities in PTSD.
|
|
| 6. |
- Dean, Kelsey R., et al.
(författare)
-
Multi-omic biomarker identification and validation for diagnosing warzone-related post-traumatic stress disorder
- 2020
-
Ingår i: Molecular Psychiatry. - : Springer Science and Business Media LLC. - 1359-4184 .- 1476-5578. ; 25:12, s. 3337-3349
-
Tidskriftsartikel (refereegranskat)abstract
- Post-traumatic stress disorder (PTSD) impacts many veterans and active duty soldiers, but diagnosis can be problematic due to biases in self-disclosure of symptoms, stigma within military populations, and limitations identifying those at risk. Prior studies suggest that PTSD may be a systemic illness, affecting not just the brain, but the entire body. Therefore, disease signals likely span multiple biological domains, including genes, proteins, cells, tissues, and organism-level physiological changes. Identification of these signals could aid in diagnostics, treatment decision-making, and risk evaluation. In the search for PTSD diagnostic biomarkers, we ascertained over one million molecular, cellular, physiological, and clinical features from three cohorts of male veterans. In a discovery cohort of 83 warzone-related PTSD cases and 82 warzone-exposed controls, we identified a set of 343 candidate biomarkers. These candidate biomarkers were selected from an integrated approach using (1) data-driven methods, including Support Vector Machine with Recursive Feature Elimination and other standard or published methodologies, and (2) hypothesis-driven approaches, using previous genetic studies for polygenic risk, or other PTSD-related literature. After reassessment of ~30% of these participants, we refined this set of markers from 343 to 28, based on their performance and ability to track changes in phenotype over time. The final diagnostic panel of 28 features was validated in an independent cohort (26 cases, 26 controls) with good performance (AUC = 0.80, 81% accuracy, 85% sensitivity, and 77% specificity). The identification and validation of this diverse diagnostic panel represents a powerful and novel approach to improve accuracy and reduce bias in diagnosing combat-related PTSD.
|
|
| 7. |
- Muhammad, Shamsuddeen Hassan, et al.
(författare)
-
SemEval-2023 Task 12 : Sentiment Analysis for African Languages (AfriSenti-SemEval)
- 2023
-
Ingår i: Proceedings of the 17th International Workshop on Semantic Evaluation (SemEval-2023). - : Association for Computational Linguistics. - 9781959429999 ; , s. 2319-2337
-
Konferensbidrag (refereegranskat)abstract
- We present the first Africentric SemEval Shared task, Sentiment Analysis for African Languages (AfriSenti-SemEval) - The dataset is available at https://github.com/afrisenti-semeval/afrisent-semeval-2023. AfriSenti-SemEval is a sentiment classification challenge in 14 African languages: Amharic, Algerian Arabic, Hausa, Igbo, Kinyarwanda, Moroccan Arabic, Mozambican Portuguese, Nigerian Pidgin, Oromo, Swahili, Tigrinya, Twi, Xitsonga, and Yorb (Muhammad et al., 2023), using data labeled with 3 sentiment classes. We present three subtasks: (1) Task A: monolingual classification, which received 44 submissions; (2) Task B: multilingual classification, which received 32 submissions; and (3) Task C: zero-shot classification, which received 34 submissions. The best performance for tasks A and B was achieved by NLNDE team with 71.31 and 75.06 weighted F1, respectively. UCAS-IIE-NLP achieved the best average score for task C with 58.15 weighted F1. We describe the various approaches adopted by the top 10 systems and their approaches.
|
|
