Getting Things Done by David Allen gave me a framework to get out of the weeds when I am overwhelmed (usually once or twice a year when I stretch myself): build a to-do list that us complete enough to stop thinking about what you have to do, if a new task take less than 5 minutes just do it right away, and then prioritise the rest.
Deep Work by Cal Newport gave me a way to think about my time management: information work is not the same as a factory line where doing the same thing at similar productivity from 9 to 5 makes sense, and it is important to dedicate long stretch of quality time to be productive (vs busy).
There are no silver bullets, but learning what worked for a group of people, testing it for myself, adapting it, and using it as needed has been helpful to me.
> If you look at the reality instead of aggregated statistics and packaged narratives, you will find that gun deaths in US very much are not of the charts in places where religiosity is off the chart. If anything, these two things are negatively correlated: the more religious a place is, the fewer gun deaths (despite no scarcity of firearms among the religious).
I don't think they are comparable. DuckDuckGo federate the results of other search engines (e.g. Bing) and add its own sauce on top of it. There is a limit to what it can do.
Google invested in the fundamental from the start and they pushed the limit of what search engines can do.
BM25 has been available for Lucene since version 4 thanks to a refactoring that allowed anyone to plug in a custom relevance score if I remember correctly . The big news is that starting version 6, BM25 is the default.
I started using ConTeXt (http://wiki.contextgarden.net/Main_Page) after many years of LaTeX. The command interfaces are more homogeneous than with LaTeX and it is quite easy to do complex layout with it.
I am currently using this toolkit and I must say that I really like it.
The main advantages of mallet over weka (the main java toolkit used in academic machine learning) for Natural Language Processing are:
- No need to map words and features to position in a feature vector yourself.
- Instances preprocessing can be defined in pipes that can be saved along the models. So no need to remember the pre-processing of data for each experiments.
- Contains algorithms for structured learning (CRF, HMM and general graphic models).
On the other hand, Mallet implements less algorithm (e.g. no Support Vector Machines to my knowledge).
In short, it is a nice toolkit to be aware of if you are planning to do Natural Language Processing.
Deep Work by Cal Newport gave me a way to think about my time management: information work is not the same as a factory line where doing the same thing at similar productivity from 9 to 5 makes sense, and it is important to dedicate long stretch of quality time to be productive (vs busy).
There are no silver bullets, but learning what worked for a group of people, testing it for myself, adapting it, and using it as needed has been helpful to me.