That’s the latest scary assessment of Twitter usage metrics, courtesy of Web security firm Barracuda Networks. The security specialist says they looked at Twitter’s 19 million accounts and determined nearly four out of five (or, 79 percent) Twitter users to be inactive, or, in Barracuda Network’s terms, they are not “true Twitter users.”

Who exactly is a “true Twitter user” then? In one of the more rigorous definitions of true Twitter usage out there, Barracuda says an active or “True” Twitter user has:

• at least 10 followers
• follows at least 10 people
• and had tweeted at least 10 times.

Pretty fair definition. No?

By this measure, that comes out to 21 percent, or just under 4 million active Twitter users. The report is pretty alarmist about the other 15 million dodgy Twitter accounts, typical for a sky-is-falling online security firm. For example, it claims that since October “one in eight [new] accounts created was deemed to be malicious, suspicious or otherwise misused and subsequently suspended.”

To be sure, Barracuda Network’s findings don’t necessarily jibe with other “State of the Twitterverse” reports out there. Going back a bit further in time, for example, Hubspot found that by the end of 2008 35 percent of Twitter users had 10 or fewer followers, and 9 percent followed nobody at all. Not great, but a far cry from the 4-out-of-5-are-zombies stat.

There are all kinds of explanations for the relatively high inactive user figures on Twitter. It’s true Twitter has gone through a rapid growth faze, then decline. It’s entirely plausible Twitter is populated by a lot of newbies still unsure how and what to Tweet. In any event, Twitter itself should better define its user base. The soon-to-launch Twitter advertising platform hinges on it.

UPDATE: This assessment of Twitter already has its detractors. One of the more interesting contrarian views out there comes from Jakob Nielsen who gives us the 90-9-1 rule to explain Twitter usage, neatly summed up in this graphic.