Increasing demand for fashion recommendation raises a lot of challenges for online shopping platforms and fashion communities. In particular, there exist two requirements for fashion out t recommendation: the Compatibility of the generated fashion out ts, and the Personalization in the recommendation process. In this paper, we demonstrate these two requirements can be satis ed via building a bridge between out t generation and recommendation. rough large data analysis, we observe that people have similar tastes in individual items and out ts. erefore, we propose a Personalized Out t Generation (POG) model, which connects user preferences regarding individual items and out ts with Transformer architecture. Extensive o ine and online experiments provide strong quantitative evidence that our method outperforms alternative methods regarding both compatibility and personalization metrics. Furthermore, we deploy POG on a platform named Dida in Alibaba to generate personalized out ts for the users of the online application iFashion.is work represents a rst step towards an industrial-scale fashion out t generation and recommendation solution, which goes beyond generating out ts based on explicit queries, or merely recommending from existing out t pools. As part of this work, we release a large-scale dataset consisting of 1.01 million out ts with rich context information, and 0.28 billion user click actions from 3.57 million users. To the best of our knowledge, this dataset is the largest, publicly available, fashion related dataset, and the rst to provide user behaviors relating to both out ts and fashion items.

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease characterized by progressive muscular weakness and atrophy. Several morphometric studies have been conducted to investigate the gray matter volume or thickness changes in ALS, whereas the cortical folding pattern remains poorly understood. In the present study, we applied a surface-based local gyrification index (LGI) from high resolution MRI data to quantify the cortical folding in matched samples of 25 ALS patients versus 25 healthy controls. Using resting-state fMRI data, we further conducted seed-based functional connectivity analysis to explore the functional correlate of the cortical folding changes. We found that ALS patients had significantly reduced LGI in right occipital cortex and that abnormality in this region associated with decreased functional connectivity in the bilateral precuneus. This set of findings was speculated to result from disturbed white matter connectivity in ALS. In the patient group, we revealed significant negative correlations between disease duration and the LGIs of a cluster in the left superior frontal gyrus, which may reflect the cognitive deterioration in ALS. In summary, our results suggest that LGI may provide a useful means to assess ALS-related neurodegeneration and to study the pathophysiology of ALS.

The human left inferior parietal lobule (LIPL) plays a pivotal role in many cognitive functions and is an important node in the default mode network (DMN). Although many previous studies have proposed different parcellation schemes for the LIPL, the detailed functional organization of the LIPL and the exact correspondence between the DMN and LIPL subregions remain unclear. Mounting evidence indicates that spontaneous fluctuations in the brain are strongly associated with cognitive performance at the behavioral level. However, whether a consistent functional topographic organization of the LIPL during rest and under task can be revealed remains unknown. Here, they used resting-state functional connectivity (RSFC) and task-related coactivation patterns separately to parcellate the LIPL and identified seven subregions. Four subregions were located in the supramarginal gyrus (SMG) and three subregions were located in the angular gyrus (AG). The subregion-specific networks and functional characterization revealed that the four anterior subregions were found to be primarily involved in sensorimotor processing, movement imagination and inhibitory control, audition perception and speech processing, and social cognition, whereas the three posterior subregions were mainly involved in episodic memory, semantic processing, and spatial cognition. The results revealed a detailed functional organization of the LIPL and suggested that the LIPL is a functionally heterogeneous area. In addition, the present study demonstrated that the functional architecture of the LIPL during rest corresponds with that found in task processing. Hum Brain Mapp 38:1659-1675, 2017. © 2017 Wiley Periodicals, Inc.

Right hemispheric dominance in tonal bilingualism is still controversial. In this study, we investigated hemispheric dominance in 30 simultaneous Bai-Mandarin tonal bilinguals and 28 Mandarin monolinguals using multimodal neuroimaging. Resting-state functional connectivity (RSFC) analysis was first performed to reveal the changes of functional connections within the language-related network. Voxel-based morphology (VBM) and tract-based spatial statistics (TBSS) analyses were then used to identify bilinguals' alterations in gray matter volume (GMV) and fractional anisotropy (FA) of white matter, respectively. RSFC analyses revealed significantly increased functional connections of the right pars-orbital part of the inferior frontal gyrus (IFG) with right caudate, right pars-opercular part of IFG, and left inferior temporal gyrus in Bai-Mandarin bilinguals compared to monolinguals. VBM and TBSS analyses further identified significantly greater GMV in right pars-triangular IFG and increased FA in right superior longitudinal fasciculus (SLF) in bilinguals than in monolinguals. Taken together, these results demonstrate the integrative role of the right IFG in tonal language processing of bilinguals. Our findings suggest that the intrinsic language network in simultaneous tonal bilinguals differs from that of monolinguals in terms of both function and structure.

No abstract