118 research outputs found
HMAR: Hierarchical Masked Attention for Multi-Behaviour Recommendation
Get PDFIn the context of recommendation systems, addressing multi-behavioral user
interactions has become vital for understanding the evolving user behavior.
Recent models utilize techniques like graph neural networks and attention
mechanisms for modeling diverse behaviors, but capturing sequential patterns in
historical interactions remains challenging. To tackle this, we introduce
Hierarchical Masked Attention for multi-behavior recommendation (HMAR).
Specifically, our approach applies masked self-attention to items of the same
behavior, followed by self-attention across all behaviors. Additionally, we
propose historical behavior indicators to encode the historical frequency of
each items behavior in the input sequence. Furthermore, the HMAR model operates
in a multi-task setting, allowing it to learn item behaviors and their
associated ranking scores concurrently. Extensive experimental results on four
real-world datasets demonstrate that our proposed model outperforms
state-of-the-art methods. Our code and datasets are available here
(https://github.com/Shereen-Elsayed/HMAR)
sj-pdf-1-cre-10.1177_02692155211036956 – Supplemental material for Comparative effectiveness study of low versus high-intensity aerobic training with resistance training in community-dwelling older men with post-COVID 19 sarcopenia: A randomized controlled trial
No full textSupplemental material, sj-pdf-1-cre-10.1177_02692155211036956 for Comparative effectiveness study of low versus high-intensity aerobic training with resistance training in community-dwelling older men with post-COVID 19 sarcopenia: A randomized controlled trial by Gopal Nambi, Walid Kamal Abdelbasset, Saud M. Alrawaili, Shereen H. Elsayed, Anju Verma, Arul Vellaiyan, Marwa M. Eid, Osama R. Aldhafian, Naif A. Nwihadh and Ayman K. Saleh in Clinical Rehabilitation</p
sj-pdf-2-cre-10.1177_02692155211036956 – Supplemental material for Comparative effectiveness study of low versus high-intensity aerobic training with resistance training in community-dwelling older men with post-COVID 19 sarcopenia: A randomized controlled trial
No full textSupplemental material, sj-pdf-2-cre-10.1177_02692155211036956 for Comparative effectiveness study of low versus high-intensity aerobic training with resistance training in community-dwelling older men with post-COVID 19 sarcopenia: A randomized controlled trial by Gopal Nambi, Walid Kamal Abdelbasset, Saud M. Alrawaili, Shereen H. Elsayed, Anju Verma, Arul Vellaiyan, Marwa M. Eid, Osama R. Aldhafian, Naif A. Nwihadh and Ayman K. Saleh in Clinical Rehabilitation</p
Gender identification for Egyptian Arabic dialect in twitter using deep learning models
No full textAlthough the number of Arabic language writers in social media is increasing, the research work targeting Author Profiling (AP) is at the initial development phase. This paper investigates Gender Identification (GI) (male or female) of authors posting Egyptian dialect tweets using Neural Networks (NN) models. Various architectures of NN are explored with extensive parameters’ selection such as simple Artificial Neural Network (ANN), Convolutional Neural Network (CNN), Long–Short Term Memory (LSTM), Convolutional Bidirectional Long-Short Term Memory (C-Bi-LSTM) and Convolutional Bidirectional Gated Recurrent Units (C-Bi-GRU) NN which is tuned for the GI problem at hand. The best acquired GI accuracy using C-Bi-GRU multichannel model is 91.37%. It is worth noting that the presence of the bidirectional layer as well as the convolutional layer in the NN models has significantly enhanced the GI accuracy
Beirut is Burning: Drag in the Creation of Queer Lebanese Identities
No full textThis thesis analyzes the role of drag performance in constructing a queer Lebanese identity and creating queer spaces and a queer community in Beirut, bound by a sectarian and communal context but also subjected to global processes of theorizing and enacting gender and sexuality. The project will thus focus upon both the subjectivities of the performers and the sociopolitical circumstances within which they operate to fully conceptualize the novelty of hybrid identities that are not located at the intersection of competing ideologies but, by constantly disrupting and challenging them, work to create new selves
Gender identification of egyptian dialect in twitter
No full textDespite the widespread of social media among all age groups in Arabic countries, the research directed towards Author Profiling (AP) is still in its early stages. This paper provides an Egyptian Dialect Gender Annotated Dataset (EDGAD) obtained from Twitter as well as a proposed text classification solution for the Gender Identification (GI) problem. The dataset consists of 70,000 tweets per gender. In text classification, a Mixed Feature Vector (MFV) with different stylometric and Egyptian Arabic Dialect (EAD) language-specific features is proposed, in addition to N-Gram Feature Vector (NFV). Ensemble weighted average is applied to the Random Forest (RF) with MFV and Logistic Regression (LR) with NFV. The achieved gender identification accuracy is 87.6%. Keywords: Text classification, Egyptian Arabic, Gender identification, Author profiling, Gender annotated datase
#747 Women and the New East.
No full textParticipants include: Begum Shereen Aziz Ahmed, Wife of the Ambassador of Pakistan to the U.S. Mrs. Hazami Fekini, Wife of the Ambassador of Libya to the U.S. Lillian T. Mowrer, Lecturer and Author of Journalist's Wife and the Indomitable John Scot
Race and Criminal Justice in Canada: An Overview
No full textCanadian Law and Society Association Annual Meeting 2014. University of Manitoba, Faculty of Law, Winnipeg, M
Laparoscopic resection and repair of caesarean scar pregnancy
Get PDFThe presented work is case series over 2 years of caesarean scar pregnancy over 2 years from January 2020 to January 2022 in Zinat Alhayat hospital of maternity in Benha city Egypt. Cases recruited from those attending Zeinat Alhayat maternity hospital in Benha and all case proved to have caesarean scar pregnancy by ultrasonography and quantitative HCG the total number of cases were 15 over a period of two years, most of patients complained about abnormal uterine bleeding in the first trimester with abnormal abdominal pain, all cases prepared for laparoscopy in Zinat Alyayat hospital in Benha and a written consent taken then with general anesthesia pelvis and abdomen explored by laparoscopy and the site of the scar opened with a hook with the aid of a traumatic grasper and then sac evacuated and the old scar resected by laparoscopic scissor. Regarding epidemiological data of patients there were no statistically significant difference in age body weight age or the amount of pain by facial analogue scale of pain. All patients saved and laparoscopy done with an average time of 45 min with no operative or postoperative complications, only one of the cases with severe bleeding required blood transfusion of 2 units of blood because HB was 7.8 g/dl, so laparoscopic treatment of caesarean scar pregnancies is a good option for patients with short operative time and good outcomes without complications.
CARCA: Context and Attribute-Aware Next-Item Recommendation via Cross-Attention
Get PDFIn sparse recommender settings, users' context and item attributes play a
crucial role in deciding which items to recommend next. Despite that, recent
works in sequential and time-aware recommendations usually either ignore both
aspects or only consider one of them, limiting their predictive performance. In
this paper, we address these limitations by proposing a context and
attribute-aware recommender model (CARCA) that can capture the dynamic nature
of the user profiles in terms of contextual features and item attributes via
dedicated multi-head self-attention blocks that extract profile-level features
and predicting item scores. Also, unlike many of the current state-of-the-art
sequential item recommendation approaches that use a simple dot-product between
the most recent item's latent features and the target items embeddings for
scoring, CARCA uses cross-attention between all profile items and the target
items to predict their final scores. This cross-attention allows CARCA to
harness the correlation between old and recent items in the user profile and
their influence on deciding which item to recommend next. Experiments on four
real-world recommender system datasets show that the proposed model
significantly outperforms all state-of-the-art models in the task of item
recommendation and achieving improvements of up to 53% in Normalized Discounted
Cumulative Gain (NDCG) and Hit-Ratio. Results also show that CARCA outperformed
several state-of-the-art dedicated image-based recommender systems by merely
utilizing image attributes extracted from a pre-trained ResNet50 in a black-box
fashion
- …
