How does DNSWL work? It's clear to me when someone report a spam, but how can I gain a higher score/trust?
Do some companies report "good" servers? Or will the score go up with time without reporting spam to them?
DNSWL usually refers to DNSWL.org but can more generally refer to the DNS-based Whitelist concept itself, which is the reverse of a DNS Blocklist (DNSBL, a.k.a. DNS Blacklist, Blackhole List, or its original name of Real-time Blackhole List, RBL). On the commercial level, there are several providers of "IP Reputation" services that assign a score to most IPs that indicate its level of danger or safeness. SenderBase is one such service.
Different services will use different methods for determining whether or not to list an IP, and much of that is going to be a trade secret. Many DNS Whitelists (and all IP Reputation systems) are automatic, but DNSWL.org requires a submission in order to kick off the process (which I think must come from the IP owner), then basically uses spam traps and spam report systems to determine how clean the given IPs are.
From the DNSWL.org website:
Your email filter should try to avoid tagging messages as spam if they come from one of the good mailservers we list, which are assigned into one of four trust levels:
TRUST LEVEL DESCRIPTION High Highly unlikely to send spam, and expected to be very fast in remediating issues. Medium Rare spam occurrences, corrected promptly. Low Occasional spam occurrences, actively corrected but less promptly. This is the default for most categories. None These are legitimate mail servers, but they may also emit spam or have other issues from time to time. This is the default for some categories (eg Email Marketing Provider).
DNSWL.org, like most services that track good reputation, determines trust by looking at (a lack of) violations over time. A high volume mail relay without spam reports will slowly gain an increased level of trust. A low volume mail relay will likely never get better than "Low" in DNSWL.org's system.