In each of these lists, we are counting the number of messages (m), senders of messages (s) and subscribers (S) over the observed period of time (t). Thus, we are computing the activity of the list (m/t) and the participation in the list (s/S %).

Next, for each mailing list, we are counting how many mailings each subscriber has sent. Then we are forming a table showing how many subscribers have sent a certain number of messages. Normalizing the columns of the latter table, we are obtaining the Histogram (number of messages vs. number of subscribers) for each mailing list. Here we might comment that subscribers who have sent 0 (or very few) messages are who are named "lurkers" in the Internet jargon.

What is quite original in our work is that we are studying threads of messages in each of the studied mailing lists: a thread is a set of messages on the same topic of discussion, i.e., a series of messages in which a message is replying the previous one. We are counting the total number messages in threads (TM) and the number of subscribers participating in threads (TS). So, we are computing the percentage of threaded messages in a list (TM/m %) and the percentage of participation in threads in a list (TS/S %). In addition, for each thread in each list, we are counting how long the thread of messages lasts.

We are also identifying the groups of countries of the participating senders in threads of messages. We are using the following groups: E = Europe, A = America, AAA = Asia, Australia, Africa, Var. = .com, .net, .org, etc. In particular, for the EuroCon-KnowFlow list, we are considering the participation from the various countries in the SOEIS project: Ch-p = Switzerland (in project), De-p = Germany (in project), Dk-p = Denmark (once in project), Gr-p = Greece (in project), It-p = Italy (in project), Nl-p = Netherlands (in project), UK-p = United Kingdom (in project).

Finally, we are sorting both the total messages and the threaded messages in a list according to thematic topics being discussed and we are computing percentages of messages in each topic. We are considering the following topics: announcements, administrative, theory, list maintenance, and miscellaneous. From the above analyses, one could draw some various conclusions on patterns of communicational dynamics emerging in each mailing list. In fact, we are producing a comparison of all the data produced from the analysis of all the examined mailing list. However, an interesting conclusion refers to the existing communicational connectivity in each mailing list, which can be inferred from the statistics of threads. In fact, the actual communicational activity in a discussion list depends on how many discussions are taking place and how many subscribers are involved in these discussions. Of course, such an activity is manifested in the statistics on threads: A communicationally active mailing list would include a high number of threads and a high participation of subscribers in these threads. On the opposite, in a communicationally inactive list most of the distributed messages would be unanswered mailings (like announcements), i.e., such a list would be poor in threads.

In this respect, it is also very interesting to understand the possible variety of people participating in threads: If they are almost the same, the percentage of participation will be low. Furthermore, an interesting point to know is whether there are particular groups of subscribers to which communication in the list are restricted. If this were the case, then the mailing list would manifest symptoms of what is called "informational Balkanization." This is why, i.e., to see this occurrence in the studied mailing lists, to each thread we are associating the groups of countries of participants in threads.