The list below includes some ideas that caught our attention. It is meant to help stimulate your thinking. You do not need to adopt an idea from this list.

In general, you will do well to poke around the web a bit. Some of the best (and peer reviewed) work in NLP appears in NLP conferences such as ACL, EMNLP, and NAACL as well as in general AI conferences such as AAAI and IJCAI. Also look for SemEval. (These acronyms should become clear if you search for them along with a year, e.g., ACL 2019.)

When you find a paper whose title catches your eye, read its abstract (which may or may not be entirely clear since abstracts generally lack examples). Read through the introduction (which should have some examples) and conclusions at first to decide whether to read the whole paper. When you read the whole paper, skip over the parts that sound incomprehensible—your goals is not to understand the paper but to get a sense for the problem and the general nature of the solution.

Think of a problem you would like to work on along with a possible approach. Subsequently, also identify some hypotheses and a means to evaluate them.

• Dealing with novel words. See https://arxiv.org/pdf/1811.03866.pdf
• Longer list
• I can't think of a natural use of simulation in language itself but if you relate language use or evolution to some element of society, then simulation can be quite appropriate. You would have to ensure the approach is realistic with respect to human languages.