May 9, 2014


Wow. Work has been super busy the past two months. It’s good in that it’s all Big Data related and very exciting. The bad part is that it’s left little time for data science projects and blogging.

But, I haven’t been completely idle. I’ve been working on a topic model to discover themes in the entries from my employer’s annual innovation contest. The goal is to determine if such a model can identify emerging innovations and latent experts within the company. The exercise has been both interesting and frustrating. I’m using the R tm and topicmodels packages which are easy to work with but sometimes slow. This is particularly painful while trying to determine the optimal number of topics. Creating a topic map from the data has also been challenging. I’ll post about these topics when I get the chance.

I’ve also been collaborating on a social network community detection project. I’m just starting to participate in this effort and I hope to post about it in the future.

Since work will likely be busy for a while, I’m going to try writing shorter posts on a more frequent basis. Also, inspired by Cal Newport’s writings on Deep Work, I plan to read one academic paper per week and post my thoughts about it.

At least, that is my plan….

Tags:  Meta