Tianyi Hao

TL;DR: A novel framework for user identification in cyber-physical space is proposed, which consists of three key steps: modeling the location distribution of each IP address; computing the co-occurrence with an inverted index to reduce the space and time cost; and a learning-to-rank tactic to fuse user's features shared in both spaces to improve the accuracy.

TL;DR: In this article, the authors proposed a method of establishing a credit investigation model, a credit determination method and corresponding apparatus thereof, which comprises the steps: collecting the visiting characteristics of a user with the known credit status as a training sample; utilizing the training sample to train a machine learning model, and obtaining a credit investigations model; collecting the user visiting characteristics, and inputting the user's visiting characteristics into the credit investigations, and finally, obtaining the credit status of the user to be evaluated.

TL;DR: Tuv is designed based on a novel combination of word2vec, negative sampling and a smoothing technique for model training, which jointly handles multi-format user data and computes user representing vectors, by taking into consideration user feature variation, self-similarity and pairwise interactions among users.

TL;DR: This paper uses a clustering method based on locations of IP addresses to map IP addresses into specific physical location distributions and proposes a novel pairwise index to reduce space cost and running time for computing the co-occurrence.

TL;DR: In this paper, the authors disclosed a method and device for determining a permission and a specific implementation way of the method comprises the following steps: acquiring historic geographic position information set in response to a received login request aiming at a target account by a request terminal; executing the following authentication steps: selecting historic geographic location information from the historic geographic positioning information set, generating a problem related to the selected historic location information, sending the problem to the request terminal, determining a frequency threshold value of sending the problems to the problem based on a condition whether answer information aiming at the problem and sent by