Detection of Topics and Construction of Search Rules on Twitter

被引：0

作者：

Martinez, Eduardo D. ^{[1
]}

Fonseca, Juan P. ^{[1
]}

Gonzalez, Victor M. ^{[1
]}

Garduno, Guillermo ^{[2
]}

Huipet, Hector H. ^{[1
]}

机构：

[1] ITAM, Mexico City 01080, DF, Mexico

[2] Sinnia, Mexico City 11560, DF, Mexico

来源：

ADVANCES IN COMPUTING, CCC 2017 | 2017年 / 735卷

关键词：

Twitter; Topic; Detection; Clustering;

D O I：

10.1007/978-3-319-66562-7_13

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

This study proposes an improvement to the Insight Centre for Data Analytics algorithm, which identifies the most relevant topics in a corpus of tweets, and allows the construction of search rules for that topic or topics, in order to build a corpus of tweets for analysis. The improvement shows above 14% improvement in Purity and other metrics, and an execution time of 10% compared to Latent Dirichlet Allocation (LDA).

引用

页码：171 / 183

页数：13

共 24 条

[1]

Adarsh M., 2015, INT J COMPUTER APPL, V128, P34, DOI [10.5120/ijca2015906553, DOI 10.5120/IJCA2015906553]

[2]

Agarwal V., 2015, INT J COMPUT APPL, V132, P13

[3]

Anjaria M., 2014, P 6 INT C COMM SYST, P1

[4]

[Anonymous], [No title captured]

[5]

[Anonymous], 2017, TASS TALLER ANALISIS

[6]

[Anonymous], 2008, Introduction to information retrieval

[7]

[Anonymous], 2011, MODERN HIERARCHICAL

[8]

[Anonymous], INT J COMPUT APPL

[9]

[Anonymous], TWITTER IN NUMBERS

[10]

[Anonymous], 2001, Snowball: a language for stemming algorithms

← 1 2 3 →