LExL: A Learning Approach for Local Expert Discovery on Twitter

Niu, Wei; Liu, Zhijiao; Caverlee, James

doi:10.1007/978-3-319-30671-1_71

Cited by 3 publications

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is mainly based on the probability statistics model, which solves the expert users' ranking by calculating the probability that users are experts. Niu et al 11,12 proposed a local expert sorting algorithm named LExL, 12 using Microsoft's famous LambdaMART 13 algorithm in "Learning to rank" 14 from four dimensions: user's own attributes, tag table, location authority, and location-based random walk to sort candidate users. 4 One is the profile-centric method, which measures the correlation degree between the profiles created for users and the queries; the other one is the document-centric method, which ranks the experts and the documents related to the query.…”

Section: Related Workmentioning

confidence: 99%

“…However, this algorithm ignores the network topology and the large and abundant content information and the cluster center in a city cannot fully reflect a user's active points. Niu et al proposed a local expert sorting algorithm named LExL, using Microsoft's famous LambdaMART algorithm in “Learning to rank” from four dimensions: user's own attributes, tag table, location authority, and location‐based random walk to sort candidate users.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Local experts finding using user comments in location‐based social networks

Cao

Yang

Cao

et al. 2019

Trans Emerging Tel Tech

View full text Add to dashboard Cite

The opinions of local experts in the location‐based social network are of great significance to the collection and dissemination of local information. In this paper, we investigated in‐depth how the user comments can be used to identify the local expert over social networks. We first illustrate the existences of potential local experts in a social network using a scored model by considering the personal profiles, comments, friend relationship, and location preferences. Then, a multidimensional model is proposed to evaluate the local expert candidates and a local expert discovery algorithm is proposed to identify local experts. Meanwhile, a scoring algorithm is proposed to train the weights in the model. Finally, an expert recommendation list can be given based on the score ranks of the candidates. Experimental results demonstrate the effectiveness of proposed model and algorithms.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Local experts finding using user comments in location‐based social networks

Cao

Yang

Cao

et al. 2019

Trans Emerging Tel Tech

View full text Add to dashboard Cite

show abstract

On Topic Aware Recommendation to Increase Popularity in Microblogging Services (Short Paper)

Litou

Kalogeraki

Gunopulos³

2016

On the Move to Meaningful Internet Systems: OTM 2016 Conferences

View full text Add to dashboard Cite

On Local Expert Discovery via Geo-Located Crowds, Queries, and Candidates

Niu¹,

Liu²,

Caverlee³

2016

ACM Trans. Spatial Algorithms Syst.

View full text Add to dashboard Cite

Local experts are critical for many location-sensitive information needs, and yet there is a research gap in our understanding of the factors impacting who is recognized as a local expert and in methods for discovering local experts. Hence, in this article, we explore a geo-spatial learning-to-rank framework for identifying local experts. Three of the key features of the proposed approach are: (i) a learning-based framework for integrating multiple user-based, content-based, list-based, and crowd-based factors impacting local expertise that leverages the fine-grained GPS coordinates of millions of social media users; (ii) a location-sensitive random walk that propagates crowd knowledge of a candidate’s expertise; and (iii) a comprehensive controlled study over AMT-labeled local experts on eight topics and in four cities. We find significant improvements of local expert finding versus two state-of-the-art alternatives, as well as evidence for the generalizability of local expert ranking models to new topics and new locations.

show abstract

LExL: A Learning Approach for Local Expert Discovery on Twitter

Cited by 3 publications

References 9 publications

Local experts finding using user comments in location‐based social networks

Local experts finding using user comments in location‐based social networks

On Topic Aware Recommendation to Increase Popularity in Microblogging Services (Short Paper)

On Local Expert Discovery via Geo-Located Crowds, Queries, and Candidates

Contact Info

Product

Resources

About