游戏评论数据挖掘

数据挖掘 机器学习 数据挖掘
2022-03-05 18:54:05

我正在尝试对游戏的评论进行文本挖掘,并希望从数据的评论示例中找到有趣的事情是

'review': ['Simple yet with great replayability. In my opinion does "zombie" hordes and team work better than left 4 dead plus has a global leveling system. Alot of down to earth "zombie" splattering fun for the whole family. Amazed this sort of FPS is so rare.',
'Amazing, Non-stop action of blowing stuff to bits, Decapitation and shooting everything you see. With a combination of action, thriller and emmersive gameplay, as well as enviromental challanges (Jump physics). This game will really put your eyes to the test, can you see the enemys before they see you? Cause their are so many!This is the second level of the killing floor, I quote bill LF4D "Son we just crossed the street" But in reality they only moved up an elevator level on genes. What has yet to come as the game is slowly realsed with thrilling and horryfing creations, But who really cares, Let\'s just blow it up, I invite you to get on the GODAMN KILLING FLOOR, LET'S SHOOT  AND GET PAID!',......
]

我有兴趣从评论数据中找到游戏的名称和游戏类型(策略、射击等)。

我尝试并创建了正则表达式,以便获取具有数据的文本(是游戏),以便我获得评论

Dying Light is a game.
Magicka is a game

我想要一个改进上述结果的建议,我的兴趣是给我发短信并找到游戏类型,如拼图、射击等

1个回答

改进结果的建议是使用NER您至少可以提取名称和类型。多搜索一下,也许有人也接受过不同类型/流派的游戏培训,所以你也可以区分这一点。