Seminar on Internet Technologies (Summer 2020): Difference between revisions

Jump to navigation Jump to search

Seminar on Internet Technologies (Summer 2020) (view source)

Revision as of 10:10, 21 February 2020

1,702 bytes added ,  21 February 2020
→‎Topics
Line 52: Line 52:
|{{Hl2}} |'''Initial Readings'''
|{{Hl2}} |'''Initial Readings'''
|-
|-
| User Profiling based on Deep Learning
| Knowledge Graph for Recommendation System
| User profiling is to infer the user's profile (e.g., age) from user-generated data, e.g., users' posts，reviews. It can provide basic data for many personalized services in e-commerce, social networks, etc. Deep learning has become the mainstream techniques in User Profiling. We need to search,read and summarize papers on google scholar from 2015 to 2020, which are discussing user profiling based on Deep Learning.
| The success of recommendation system makes it prevalent in Web applications, ranging from search engines, E-commerce, to social media sites and news portals.To predict user preference from the key (and widely available) source of user behavior data, much research effort has been devoted to collaborative filtering (CF) [12, 13, 32]. Despite its effectiveness and universality, CF methods suffer from the inability of modeling side information [30, 31], such as item attributes, user profiles, and contexts, thus perform poorly in sparse situations where users and items have few interactions.To address the limitation of CF models, a solution is to take the graph of item side information, aka. knowledge graph into account to construct the predictive model.
| Interested in this topic, willing to follow the advisor's guidance, patience and time for reading multiple papers
| Interested in this topic, patience and time for reading and concluding multiple papers.
| [Shichang Ding,sding@gwdg.de]
| [Shichang Ding,sding@gwdg.de]
| [https://repository.kaust.edu.sa/bitstream/handle/10754/628781/p1764-liang.pdf?sequence=1&isAllowed=y]
| [https://www.google.com/search?q=kgat&oq=kgat&aqs=chrome..69i57j0l7.791j0j4&sourceid=chrome&ie=UTF-8]
|-
|-
| Empirical study for QUIC Protocol
| Empirical study for QUIC Protocol
Line 64: Line 64:
| [https://ieeexplore.ieee.org/abstract/document/8638062]
| [https://ieeexplore.ieee.org/abstract/document/8638062]
|-
|-
|  
| Failure recovery from the breakpoint in service function chain
|  
| As we all know, if the packets are dropped in network, we need to retransmit them from the sender. However, in service function chain, failure links or nodes may drop packets that have already been processed by upstream NFs, retransmission from the sender may result in wasted work in the service chain. If we use SRv6 to steer traffic along with SFC, we could easily know the IP address of upstream NF, then we can leverages this information to realize in-network recovery. This project focuses on realizing in-network recovery with SRv6.
|  
| The student should know the basic knowledge about TCP/IP, network simulation
|  
| [Bangbang Ren, bangbang.ren@cs.uni-goettingen.de]
|
|-
| Learning Combinatorial Optimization Algorithms over Graphs
| There are many NP-hard problems about graph. However, these NP-hard problems cannot be soloved fast by optimization solver. Approximation algorithms could solve them fast in the cost of sacrificing the accuracy. Recently, some algorithms based on machine learning have been proposed to solve these NP-hard problems in the manner of end-to-end. After reproducing one classical paper, the student is required to find solution for a new assignment problem
| The student should be familiar with machine learning and Integer linear programming
| [Bangbang Ren, bangbang.ren@cs.uni-goettingen.de]
|-
|}
|}


Sding
308

edits

Retrieved from "https://wiki.net.informatik.uni-goettingen.de/wiki/Special:MobileDiff/6426"

Navigation menu