Yes there are laws about spam: here's for the US, here are some links about Europe spam laws.
I'm pretty sure that in France (not sure about other countries), unsolicited commercial emails are illegal if sent to a person BUT commercial prospection IS legal so if the email addresses are professional that would be OK.
Anyway, in my opinion the ethic problem ...
Pew Research Center has a data site with downloads for many raw datasets from surveys and polling. Requires registration and license is for non-commercial use.
Track key national, political, economic and demographic trends over time using our regularly updated charts and explore further by downloading our data sets, trying our data interactives or viewing ...
Not a direct answer to the legal question, but...
If you send any unsolicited email then to prevent your sending address/domain being marked as a spammer, you should use an emailing service. Once you are marked as a spammer, which is based on people putting your email in the spam folder, then very few of your emails will get through to inboxes.
You can do sentiment analysis on social media feeds, but I suspect there is an inherent overlap (bias) between social media users (e.g. Twitter) and people who are pro-Uber.
To do so with Twitter, you can use the API to search terms related to Uber with the following constraints:
+ located in France (bounding box)
+ tweet text in the French language
I would consider asking questions similar to the data collected in the IPEDS surveys by the US Department of Education. Each year, every accredited university/college and trade school fills this survey out. It includes a large number of questions, including: tuition, books, housing on/off campus, degrees offered, size of freshman class, student population by ...
Pew Research has a dataset that was used for this article:
A survey of LGBT Americans
You can download the data here, after registering.
Findings in this report are based on two main data sources:
This report is based primarily on a Pew Research Center survey of the LGBT population conducted April 11-29, 2013, among a nationally ...
Trustworthiness is a very vague term and I don't know what kind of trustworthiness you are interested in. I will give you two examples I came across in the last few years.
We can use data to test if can trust other members of the group. For example, are civil servants upholding the law and doing their work correctly. In this blog, Ben Wellington showed ...
Search for "470220" in both and you'll see UCC defined, and 2012 is clearly different from 2013.
I'll bet the why are they different is buried in the documentation too, although I'm certainly to lazy to pick through it.