Here you can discuss any topics related to data mining and big data
- Posts: 69
- Joined: Tue Apr 05, 2022 4:48 pm
I would like to ask what are the current hot topics in data mining and data science?
- Site Admin
- Posts: 121
- Joined: Tue Apr 05, 2022 12:47 am
- Location: China
Personally, I see a lot of possibilities on working around complex data types like graphs, sequences, dynamic graphs, data streams, heterogeneous data, etc.
- Posts: 39
- Joined: Sun Apr 10, 2022 12:17 am
At my institute, almost everyone is doing NLP (natural language processing) and image/video processing these days. I find this depressing. There are many other things to do research on.
- Posts: 52
- Joined: Tue Apr 05, 2022 8:32 am
Pattern mining is an interesting topic in data science. It consists of the discovery of frequent sets, of which the support (number of occurrences of the pattern) is above a given threshold. In many cases, support is inversely correlated to the size of the patterns, so that the most interesting patterns have small support, often under 1% of the data. It has been shown that finding frequent sets is NP-hard and known to be in the worst case even hard to approximate. Also, if the data is massive or given as a stream, classical approaches are often no longer suitable, due to the size of the dataset or the sheer number of sequential scans. Developing new algorithms for pattern mining tasks is an interesting research topic with many opportunities.