T R A C K       P A P E R
Close

Login Panel

Close tab

Password Meter

Volume 09 Issue 10 (October 2022)

S.No. Title & Authors Page No View
1

Title : Review on the Research Progress of Recommendation System Algorithms

Authors : Xiaoqi Wang

Click Here For Abstract

Download Certificate
Abstract :

With the wide application of mobile Internet and smart phone, geographical social network has become an indispensable social information platform in People's Daily life. Since human is a group with typical social attributes, group activities are the state portrayal of People's Daily life and work in the real world. Therefore, in this new social information platform, the recommendation service requirements related to groups and group activities are becoming more and more important tasks in the current recommendation system. Although a large number of studies have been conducted to solve the problem of group recommendation oriented to group activities, there are still some problems in the research of group recommendation :(1) cold start  (2) data sparsity. In this paper, the research on popular recommendation algorithms is systematically analyzed and sorted out. Mainly from the current existing research methods and the advantages and disadvantages of the current scheme to summarize; Finally, the characteristics of the current algorithm and the prospect of the future are summarized.

1-3
2

Title : Blockchain-Based Traceability Study of Chinese Herbal Medicine

Authors : Jia Zixuan

Click Here For Abstract

Download Certificate
Abstract :

Chinese medicine has a pivotal position in China and the world, and Chinese herbal medicine, as an essential part of it, naturally attracts much attention. The monitoring and management of the supply chain of Chinese herbal medicine is an important link to maintain the healthy development of the Chinese herbal medicine industry. At the present stage, the supply chain supervision of Chinese herbal medicines still has problems such as information asymmetry and difficulty in traceability and accountability, which cannot guarantee the quality and safety of Chinese herbal medicines. In order to improve the quality of safety supervision of Chinese herbal medicines by relevant departments and improve the operational efficiency of the Chinese herbal medicine supply chain, we analyzed the factors affecting the traceability of Chinese herbal medicine supply chain in China and constructed a traceability system of Chinese herbal medicine supply chain using the process trustworthiness and decentralization features of blockchain to provide solutions for the traceability of Chinese herbal medicine quality and safety.

4-6
3

Title : Caching Mechanism of Decentralized Social Networks

Authors : Chenle Xiong

Click Here For Abstract

Download Certificate
Abstract :

With the rapid development of the Internet, the social traffic in the network is growing explosively. The cache pressure faced by social networks is increasing. However, traditional social networks are centralized. The cache content in the network is stored in the central server or base station, while the cache content of decentralized social networks is scattered in each node of the network. Therefore, this paper analyzes the cache mechanism and cache replacement strategy of decentralized social networks, which are completely different from centralized social networks.

7-9
4

Title : Big Data Analysis Course Case Survey

Authors : Xiaoqi Wang, Ze Wang

Click Here For Abstract

Download Certificate
Abstract :

In the era of big data, the development of location sensing technology such as mobile communication and sensing devices has formed location big data, which has brought huge benefits to people's life, business operation methods and scientific research. Due to the diverse uses of location big data and the overlapping redundancy of contents, the classical privacy protection methods based on "informed consent" and anonymity can not fully protect user privacy. The privacy protection technology of location big data measures the user's location privacy and protects the user's sensitive information in the sense of information theory. This paper introduces the concept of location big data and the privacy threats of location big data, summarizes the unified metric-based attack model for location big data privacy, and summarizes the existing research achievements in the field of location big data privacy protection. The user's knowledge of text content and online behavior can be used to carry out reasoning attacks on users. Social relational reasoning and attribute reasoning are two basic attacks on user privacy in social networks. The research on the protection mechanisms and methods of inference attacks is also increasing. This paper classifies and summarizes the research and literature related to privacy inference and protection techniques, and finally discusses and looks into the future.

10-14
5

Title : Overview of Teaching Cases of "Big Data Privacy Protection" Oriented to Practical Ability Training

Authors : Kaiyue Hu, Wenju Liu, Ze Wang

Click Here For Abstract

Download Certificate
Abstract :

Under the background of the new era, the society has put forward higher requirements for computer professionals. The traditional professional practice and innovation teaching link is faced with many problems, which is difficult to meet the needs of the society for innovative and comprehensive talents. In order to promote big data teaching reform, promote the integration of scientific research and teaching, improve students' practical ability and innovation ability, and meet the requirements of the new era of professional talent training goals, this paper analyzes and introduces the location of big data in big data. Location big data not only makes people's lives more convenient, but also brings the risk of personal sensitive information leakage. This paper summarizes the location privacy protection methods and trajectory privacy protection methods, introduces two privacy protection models, and provides a certain basis for the reform and development of space-time big data in big data courses.

15-18
6

Title : Design and application of Teaching Case of Authentication and Data Security Mechanism Based on Blockchain

Authors : Fang Li, Wenju Liu, Ze Wang

Click Here For Abstract

Download Certificate
Abstract :

The Internet provides us with rich data and services. Identity authentication and resource access authorization are key technologies in the field of network services and management. To realize the transmission of multi-domain identity information with only one cross-domain request process, we make full use of the characteristics of distributed consistency, combine the teaching experience of cross-domain identity authentication and data security in recent years, we propose a certificateless cross-domain identity authentication scheme based on master-slave chain. The scheme provides reference for the application teaching reform of data security by discussing the case design of cross-domain identity authentication.

19-23
7

Title : Prompt Learning Classification Task Based on Improved Answer Space

Authors : Jiaxi Yang, Baoshan Sun

Click Here For Abstract

Download Certificate
Abstract :

In order to solve the problem of single label word selection in prompt learning, this paper proposes a method of Probabilistic Answer Set. Add the prefixed space vector in front of the input text sentence as a template for prompt learning. Constructing an answer space set for each label word, and the probability that the text belongs to this category is obtained in the form of probability of the answer set. The experimental results show that on the THUCNews news classification datasets, the algorithm can improve the accuracy and F1-score by 1 percentage points compared with the common classification model. It can improve by 0.5 percentage points compared to the fixed prompt learning mode. It shows that the Probabilistic Answer Set algorithm proposed in this paper is better than other models significantly in the text classification task.

24-27
8

Title : Group Photo Realisation: A Survey of Multi person Group Photo Image Generation

Authors : Hongming Lu

Click Here For Abstract

Download Certificate
Abstract :

Group photography is a common image in life, but when you take a group photo, there will always be someone who can't be there for a variety of reasons, resulting in a group photo always with regret. Especially in recent years, because of the epidemic, many people at the gathering could not be reached. As a result, the research on the generation of group photo images of people becomes more and more important. The purpose of this study is to perfectly embed a single person image into a multi-person photo image to make it harmonious and realistic. Previous people's image compositions focused more on the fusion of characters and backgrounds, with little regard to the interaction between people and other people in the group photo. Now, it is more desirable to have a good fusion with multi-person photo images. In this survey, we summarize the methods of the above research directions, and discuss the difficulties and future directions of future group photo character image generation to promote future research of group photo realization.

28-30
9

Title : Chinese Painting Synthesis Based on Deep Learning: A Review

Authors : Mengxue Wang

Click Here For Abstract

Download Certificate
Abstract :

With the development of deep learning, various deep learning techniques have achieved impressive results in various image synthesis tasks. They have become a hot spot in computer vision research due to their impressive performance in various applications. Deep learning of the Chinese painting generation is one of them. This paper introduces the research progress of deep learning in the generation of Chinese paintings in recent years, mainly including the generation of Chinese paintings based on style transfer and the generation of Chinese paintings based on sketches. We analyze and summarize the datasets, methods, and pros and cons used in these applications and give an outlook on the work after the generation of Chinese paintings.

31-33
10

Title : Approximation of Geodesic Distances by Autoencoder Networks

Authors : Jiaxi Yang

Click Here For Abstract

Download Certificate
Abstract :

Autoencoder network is a kind of network structure in current deep neural network. In recent years, autoencoder networks have been used in the tasks of dimensionality reduction of various data and feature extraction of complex things. Geodesic distance describes the local shortest path between two points on a spatial manifold, which is an essential measure of the manifold.This paper uses the Swiss roll function as the main research function, combines the auto-encoding network with the geodesic distance, uses the auto-encoding network to input the coordinates of the points in the space as input values     into the auto-encoding network, and extracts and trains the features of the point coordinates. A new set of high-dimensional coordinates is obtained, so that the Euclidean space distance obtained from this set of high-dimensional coordinates can be closer to the geodesic distance in the actual space. The research content of this paper mainly includes the following three parts:

  1. The development, concept and theory of auto-encoding network are introduced, and the principle of auto-encoding training is briefly described.
  2. The Floyd algorithm is used to solve the geodesic distance between two points of the Swiss roll function, and the approximate geodesic distance is obtained.

Use the auto-encoding network to train and test the coordinates of the points in the Swiss roll function to obtain a new set of coordinates, calculate the Euclidean space distance of the new set of coordinates, and use this set of Euclidean distances with the geodesic distance measured by Floyd Make a comparative estimate and calculate the error.

34-37
11

Title : Design of "Computer Network" Innovative Teaching Method for New Engineering First-Class Talents Training

Authors : Ze Wang, Nan Song, Jing Guo

Click Here For Abstract

Download Certificate
Abstract :

T n accordance with the current development of new engineering and new infrastructure strategy for IT personnel training demand, according to the characteristics of computer network course theory and practice of equal importance to put forward the combination of online and offline hybrid teaching and online flipped classroom teaching model. Through offline teaching mixed with the online lesson "flip" teaching model to help students combining theory study and practical engineering problems, to cultivate the students' divergent thinking ability, stimulate interest in collaborative learning, lead the students into active and creative learning situations, and to strengthen the construction of political education course in the learning process, realize the unity of knowledge and value guidance

38-41
12

Title : Livelihood Rehabilitation of Project Affected People Caused by NN2 Hydropower Project in Laos

Authors : Saychai Syladeth

Click Here For Abstract

Download Certificate
Abstract :

The study aimed to examine characteristic impacts and new livelihood rehabilitation activities among PAP caused by NN2 hydropower project in Laos. The results of study revealed that, the reasons behind construction of NN2 project were not only for renewable and energy supply, but also included to eradicate insecurity and poor living conditions among rural people. The project had agrarian livelihood impacts among rural villages. Those new livelihood activities were rehabilitated by compensating lost properties, providing social services, establishing vocational groups, providing agricultural area and inputs, practicing on-farm and off-farm activities, increasing annual income, reducing poverty, implementing MDGs and enhancing capacities.

42-49
13

Title : Dielectric Study of H-Mordenite

Authors : Dr.V.P.Deshpande

Click Here For Abstract

Download Certificate
Abstract :

Zeolite Mordenite belongs to Group VII were collected near Ellora Ajanta belt. characterization was made using XRD,IR at NCL Pune.Dielectric study was made using LCR Bridge. Pellets of H-Mordenite were prepared.Variation of dielectric constant, dielectric loss, dielectric conductivity and relax session time were measured from 20Hz to 20KHz .Results were compared.

50-53