Since it is thesis proposal season for the juniors, here are some projects that I have kicking around my brain in no particular order. Many of them relate to GIS, Digital Anthropology, Computational Anthropology, or Computer Science, but there is a History and a Philosophy project in here too. Since I am not a faculty member I can't advise your capstone but I can read over drafts, provide feedback and help point you in the right direction:
Build an ArcMap compatible geocoder for IP addresses that does not reply on a closed source database. (Working on building the database could be a project within itself). Right now, you have to geocode with a propitiatory python library from a proprietary database. It would be nice to be able to do this with FOSS software and a FOSS database.
OpenAnthroStat needs a MDS module and a more efficient way of handing imported data. The fact that AnthroPac is still the dominate way of dealing with freelist data is saddening. For OAS to have a chance at breaking the dominance of AnthroPac it needs a way of generating Metric and Non-Metric MDS plots based off of the data. I would love to be able to do this without having numby as a dependency. (The goal is for this to be easier to use than AP. Each dependency makes it harder and harder to use in the field).
Study the biases in the algorithms that run our lives. I have a dataset of edits to the English language Wikipedia coded with a probability that it is a good edit and the probability that it is vandalism. Aaron Halfaker at the Wikimedia Foundation and myself are interested in geographic biases in the model used to judge edits. It would be interesting to break the data down by country, dominate language in country, and ISP type (residential, commercial, educational, university) to see if some group gets unfairly flagged. A portion of this might look into the elimination of biases against new accounts.
Wikidata. I don't feel 100% ok with publishing my thesis results until on the ground interviews are done with some of the people involved in the creation of the community. What to they think about the creation of the community. My work can be used to frame your questions better.
Other Wikimedia Projects. Wikisource and Wiktionary have seen very little love from researchers. What makes these projects tick? Why do people choose to maintain a dictionary or a repository of open source texts? There is an English language version of both and you can ground your work in previous research about Wikipedia.
Other Open Projects. Mozilla and Debian both have large communities that operate on a hybrid volunteer-staff model. How to volunteers and professionals coexist in a community?
The Great Hacker War of 1990 to 1991 is shrouded in mystery. The Wikipedia page reflects the sorry state of the scholarship of the event. Phiber Optik and other participants claim that it never happened. What it really needs is a talented historian to sort out the facts from the storytelling. (This is really a potential book along the lines of The Hacker Crackdown by Bruce Sterling and This Machine Kills Secrets by Andy Greenberg.)