Multi-interest User Profiling in Short Text Microblogs

Wandabwa, Herman Masindano; Naeem, M. Asif; Mirza, Farhaan; Pears, Russel; Nguyen, Andy

doi:10.1007/978-3-030-64823-7_15

Cited by 2 publications

(1 citation statement)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Classical collaborative filtering and matrix factorization methods do not naturally produce multiple user embeddings, and so do sequential models like RNNs and attentionbased models. To discover multiple interests from user engagement history, heuristic methods [16,45] and unsupervised learning methods like clustering [25,33] and community mining [34,44] has been adopted. Besides, researchers have made efforts to modify the existing neural networks to produce multiple results, for instance, the capsule network [1,19,28] and multi-head attention models [1,20,47].…”

Section: Related Workmentioning

confidence: 99%

Everyone's Preference Changes Differently: Weighted Multi-Interest Retrieval Model

Shi¹,

Gu²,

Zhou³

et al. 2022

Preprint

View full text Add to dashboard Cite

User embeddings (vectorized representations of a user) are essential in recommendation systems. Numerous approaches have been proposed to construct a representation for the user in order to find similar items for retrieval tasks, and they have been proven effective in industrial recommendation systems as well. Recently people have discovered the power of using multiple embeddings to represent a user, with the hope that each embedding represents the user's interest in a certain topic. With multi-interest representation, it's important to model the user's preference over the different topics and how the preference change with time. However, existing approaches either fail to estimate the user's affinity to each interest or unreasonably assume every interest of every user fades with an equal rate with time, thus hurting the recall of candidate retrieval. In this paper, we propose the Multi-Interest Preference (MIP) model, an approach that not only produces multi-interest for users by using the user's sequential engagement more effectively but also automatically learns a set of weights to represent the preference over each embedding so that the candidates can be retrieved from each interest proportionally. Extensive experiments have been done on various industrial-scale datasets to demonstrate the effectiveness of our approach. 1

show abstract

Section: Related Workmentioning

confidence: 99%

Everyone's Preference Changes Differently: Weighted Multi-Interest Retrieval Model

Shi¹,

Gu²,

Zhou³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

Machine Learning for Browser Privacy

Tan,

Logeswaran

2023

Emerging Technologies for Digital Infrastructure Development

View full text Add to dashboard Cite

Online privacy is an Internet user’s control of how much personal information is shared with a third party. Unfortunately, some third parties, such as data brokers, collect user data without permission to resell the data to other parties. Browser tracking allows each Internet user to be uniquely identified, and in-depth user profiles are built. Browser fingerprinting is one of the most effective methods of browser tracking. It uniquely identifies each user through their devices’ configuration, even for users using the same device models. Using Virtual Private Networks, the Tor browser and specific browser extensions as a countermeasure against browser fingerprinting are not widespread, so it often results in a compromised user experience. Researchers have proposed various classification machine learning approaches to improve browser privacy; some focus on recognising and blocking advertisements and website scripts that track users. In contrast, others identify potential vulnerabilities in browser security configurations. There is a need for more research in machine learning, especially natural language processing, to enhance browser privacy.

show abstract

Multi-interest User Profiling in Short Text Microblogs

Cited by 2 publications

References 25 publications

Everyone's Preference Changes Differently: Weighted Multi-Interest Retrieval Model

Everyone's Preference Changes Differently: Weighted Multi-Interest Retrieval Model

Machine Learning for Browser Privacy

Contact Info

Product

Resources

About