ashora/SocialListening
依据香港中文大学设计的规则系统,先用小样本评论建立初始关键词库,再结合18种句式逐条匹配评论,能够快速准确地识别评论对象及情感极性。经多次迭代优化关键词库后,达到较高准确率的基础上,使用Tableau进一步分析数据,识别出客户集中关注的商品属性、普遍好评差评的商品属性;通过与同类商品竞品分析,识别出商品的优缺点,画出商品属性特征风险区、改进区、保持区和低人气区分布图;最终撰写分析报告,供电商经营参考。
Python
Stargazers
- 1234567henggelolikonChina GuangDong DongGuan
- Arddddd
- BanBan13
- chenglansky
- chenjunup
- CodeOfGod
- CZU32BUPT
- DaiYingLuo
- dgodance
- EikotheRookie
- Esamalz
- ggbond123456
- hsdb
- jaylinling
- KGLEE666
- leedaihungGuangZhou, China
- lg8897203
- liaozp1
- Lilysc
- LyonJiang
- meteFANS
- mylv1222
- Orion-wycSoutheast University / Huawei HiSilicon
- putaodoudou
- rinranx
- ScottishFold007Shanghai
- simona081
- SnailDMShenZhen
- ssxjss
- taotao033Zheng Zhou, China
- Vickzhang
- wgb128
- xiaodongguadadonggua
- yuanjie-aiXiaoMi
- zhiyulee-RUC
- zhuyuuyuhzDataKnow