词语夫联矩阵
Abstract:
Micro-blog data has the characteristic of real-time,volume,short-text,and noise-rich.So it is a challenge for the traditional topic detection technology.A novel micro-blog topic detection algorithm based on topic model was proposed.Firstly,the micro-blog data was expressed as text word matrix and word relation matrix.The topic word was extracted from the two vectors.Secondly,the topic model was obtained with clustering.Finally,the topic detection of micro-blog was obtained by clustering text and topic model.Experimental results show that the algorithm proposed can effectively detection the text topic,and with the best parameter group of precision,recall rate,F,and the value F is about 95%.
Key words:
topic detection,
topic model,
text word matrix,
word relation matrix
微博话题及其数量"
|
序号
|
话题
|
数量
|
|
1
|
#北京楼顶别墅主人#
|
386
|
|
2
|
#斯诺登棱镜事件#
|
131
|
|
3
|
#李天一案件#
|
356
|
|
4
|
#郑钧刘芸马尔代夫婚礼#
|
138
|
|
5
|
#姚贝娜vs林育群#
|
143
|
|
6
|
#七夕，一起看英仙座流星雨#
|
233
|
|
7
|
#光大证券乌龙事件#
|
480
|
|
8
|
#埃及清场#
|
153
|
|
9
|
#林丹vs李宗伟#
|
350
|
|
10
|
#台风尤特#
|
609
|
|
11
|
#泰国白龙王病逝#
|
160
|
|
12
|
#中国好外婆#
|
109
|
|
13
|
#昌都地震#
|
120
|
|
14
|
#新疆吐鲁番地表温度70℃火焰山烤鸡蛋生意红火#
|
87
|
不同参数组合条件下各类话题的准确率、召回率和F值"
|
参数话题
|
α=1，β=1，γ=1
|
|
α=3，β2，γ=5
|
|
α=5，β=2，γ=2
|
|
P
|
R
|
F
|
P
|
R
|
F
|
P
|
R
|
F
|
|
1
|
24.55
|
56.48
|
34.22
|
|
100.00
|
56.21
|
71.97
|
|
98.97
|
100.00
|
99.48
|
|
2
|
11.27
|
100.00
|
20.26
|
|
95.42
|
95.41
|
95.42
|
|
99.20
|
94.66
|
96.87
|
|
3
|
14.64
|
36.51
|
20.90
|
|
97.79
|
87.07
|
92.12
|
|
98.72
|
87.08
|
92.54
|
|
4
|
8.78
|
56.52
|
15.20
|
|
100.00
|
92.75
|
96.24
|
|
100.00
|
92.75
|
96.24
|
|
5
|
35.46
|
69.93
|
47.06
|
|
97.10
|
93.71
|
95.37
|
|
99.24
|
91.61
|
95.27
|
|
6
|
88.78
|
37.33
|
52.57
|
|
75.57
|
99.57
|
85.92
|
|
82.01
|
97.85
|
89.24
|
|
7
|
41.22
|
99.79
|
58.34
|
|
26.42
|
56.04
|
35.91
|
|
99.35
|
95.83
|
97.56
|
|
8
|
6.08
|
35.29
|
10.37
|
|
84.70
|
94.12
|
89.16
|
|
96.57
|
92.16
|
94.31
|
|
9
|
29.05
|
73.71
|
41.68
|
|
20.82
|
60.57
|
30.99
|
|
99.69
|
94.57
|
97.07
|
|
10
|
60.23
|
75.86
|
67.15
|
|
29.37
|
49.09
|
36.75
|
|
98.22
|
99.67
|
98.94
|
|
11
|
9.47
|
68.75
|
16.64
|
|
95.18
|
98.75
|
96.93
|
|
95.18
|
98.75
|
96.93
|
|
12
|
5.51
|
44.95
|
9.82
|
|
100.00
|
52.29
|
68.67
|
|
99.03
|
94.50
|
96.71
|
|
13
|
6.12
|
39.17
|
10.69
|
|
96.46
|
90.83
|
93.56
|
|
100.00
|
94.17
|
97.00
|
|
14
|
6.64
|
58.62
|
11.94
|
|
100.00
|
88.51
|
93.90
|
|
100.00
|
87.36
|
93.25
|
各个话题类别的主题词"
|
话题
|
主题词
|
|
#北京楼顶别墅主人#
|
北京、别墅、楼顶、拆除、假山、张必清、顶盖、教授、山庄、小区
|
|
#斯诺登棱镜事件#
|
美国、斯诺登、棱镜
|
|
#李天一案件#
|
李天一、书法、酒吧、梦鸽、律师
|
|
#郑钧刘芸马尔代夫婚礼#
|
婚礼、马尔代夫、郑钧、刘芸
|
|
#姚贝娜vs林育群#
|
姚贝娜、声音、林育群
|
|
#七夕，一起看英仙座流星雨#
|
流星雨、流星、看到
|
|
#光大证券乌龙事件#
|
光大、证券、乌龙、交易、期货、市场、暴涨、股市
|
|
#埃及清场#
|
埃及、行动
|
|
#林丹vs李宗伟#
|
李宗伟、林丹、冠军、羽毛球、决赛、世界、对手、比赛
|
|
#台风尤特#
|
台风、尤特、强台风、登陆、中心
|
|
#泰国白龙王病逝#
|
龙王、泰国
|
|
#中国好外婆#
|
外婆、外孙、同志
|
|
#昌都地震#
|
地震、昌都、西藏
|
|
#新疆吐鲁番地表温度70℃火焰山烤鸡蛋生意红火#
|
鸡蛋、火焰山、地表
|
