This paper surveys and analyzes the state of art agent-mediated electronic commerce (e-commerce), concentrating particularly on business-to-consumer (B2C) business-to-business (B2B) aspects. From consumer buying behavior perspective, agents are being used in following activities: need identification, product brokering, buyer coalition formation, merchant negotiation. The roles B2B e-commerce discussed through transaction model that identifies as employed partnership Having identified for B2C...

Collaborative filtering (CF) is an important and popular technology for recommender systems. However, current CF methods suffer from such problems as data sparsity, recommendation inaccuracy, big-error in predictions. In this paper, we borrow ideas of object typicality cognitive psychology propose a novel typicality-based collaborative method named TyCo. A distinct feature that it finds "neighbors" users based on user degrees groups (instead the corated items users, or common items,...

Abstract This study provides a comprehensive evaluation of the efficiency Large Language Models (LLMs) in performing diverse language understanding and generation tasks. Through systematic comparison open-source models including GPT-Neo, Bloom, FLAN-T5, Mistral-7B, research explores their performance across widely recognized benchmarks such as GLUE, SuperGLUE, LAMBADA, SQuAD. Our findings reveal significant variations model accuracy, computational efficiency, scalability, adaptability,...

Changmeng Zheng, Yi Cai, Jingyun Xu, Ho-fung Leung, Guandong Xu. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and 9th International Joint (EMNLP-IJCNLP). 2019.

Visual contexts often help to recognize named entities more precisely in short texts such as tweets or snapchat. For example, one can identify "Charlie'' a name of dog according the user posts. Previous works on multimodal entity recognition ignore corresponding relations visual objects and entities. are considered fine-grained image representations. sentence with multiple types, relevant be utilized capture different information. In this paper, we propose neural network which combines...

Visual question answering aims to answer the natural language about a given image. Existing graph-based methods only focus on relations between objects in an image and neglect importance of syntactic dependency words question. To simultaneously capture question, we propose novel dual channel graph convolutional network (DC-GCN) for better combining visual textual advantages. The DC-GCN model consists three parts: I-GCN module image, Q-GCN attention alignment align representations...

Increasingly, many systems are being conceptualized, designed, and implemented as marketplaces in which autonomous software entities (agents) trade services.These services can be commodities e-commerce applications or data knowledge information economies.In of these cases, there both multiple agents that looking to procure sell at any one time.Such termed continuous double auctions (CDAs).Against this background, paper develops new algorithms buyer seller use participate CDAs.These employ...

The task of image difference captioning aims at locating changed objects in similar pairs and describing the with natural language. key challenges this are to comprehend context sufficiently locate accurately presence viewpoint change. Previous studies focus on pixel-level features, neglecting rich explicit features an pair which beneficial generate a fine-grained caption. Additionally, existing generative models suffer from differences interference To address these issues, we propose...

The recent global outbreak of Severe Acute Respiratory Syndrome has aroused public concern on environmental health and hygiene. Develops a practical assessment scheme for assessing the hygiene performance apartment buildings in Hong Kong. involves hierarchy building factors that have bearing qualities, thus occupants’ health. Proposes an index method to integrate outcomes into simple user‐ friendly indicator consumption. can inform risk different facilitate owners, developers, government...

Term weighting schemes have been widely used in information retrieval and text categorization models. In this paper, we first investigate into the limitations of several state-of-the-art term context tasks. Considering that category-specific terms are more useful to discriminate different categories, these tend smaller entropy with respect then explore relationship between a term's discriminating power its set categories. To end, propose two entropy-based (i.e., tf.dc tf.bdc) which measure...

Most existing visual question answering (VQA) models strongly rely on language bias to answer questions, i.e., they always tend fit question-answer pairs the train split and perform poorly test spilt when distributions are different. This behavior makes them hard be applied in real scenarios. To reduce biases, previous studies mainly integrate modules overcome priors (ensemble-based methods) or generate additional training data balance dataset biases (data-balanced methods). However, all...

Multi-item negotiations surround our daily life and usually involve two parties that share common or conflicting interests. Effective automated negotiation techniques should enable the agents to adaptively adjust their behaviors depending on characteristics of negotiating partners scenarios. This is complicated by fact are unwilling reveal information (strategies preferences) avoid being exploited during negotiation. In this paper, we propose an adaptive strategy, called ABiNeS, which can...