| Source | # of sentences | Average logarithmic rank |
|---|---|---|
| welt.de vom 17.06.2006 | 13 | 5.79 |
| berlinonline.de vom 01.04.2006 | 12 | 5.80 |
| welt.de vom 01.07.2006 | 11 | 5.92 |
| welt.de vom 07.03.2006 | 11 | 6.01 |
| welt.de vom 04.03.2006 | 11 | 6.02 |
| welt.de vom 06.02.2006 | 11 | 6.06 |
| welt.de vom 22.02.2006 | 11 | 6.06 |
| welt.de vom 06.04.2006 | 11 | 6.24 |
| welt.de vom 21.02.2006 | 11 | 6.35 |
| Source | # of sentences | Average logarithmic rank |
|---|---|---|
| welt.de vom 21.02.2006 | 11 | 6.35 |
| welt.de vom 06.04.2006 | 11 | 6.24 |
| welt.de vom 06.02.2006 | 11 | 6.06 |
| welt.de vom 22.02.2006 | 11 | 6.06 |
| welt.de vom 04.03.2006 | 11 | 6.02 |
| welt.de vom 07.03.2006 | 11 | 6.01 |
| welt.de vom 01.07.2006 | 11 | 5.92 |
| berlinonline.de vom 01.04.2006 | 12 | 5.80 |
| welt.de vom 17.06.2006 | 13 | 5.79 |
In this subsection we replace average word length by average logarithmic word rank. The logarithm of the word rank is taken because we want to punish words of high ranks only moderately.
First table:
select source, count(distinct i_s.s_id) as cnt_s, round(avg(log(w.w_id-100)),2) as av from sources so, inv_so i_s, inv_w i, words w where so.so_id=i_s.so_id and i_s.s_id=i.s_id and i.w_id=w.w_id and w.w_id>100 group by source having cnt_s>10 order by av LIMIT 30;
6.4.2.1 Average word length for different sources
6.4.2.3 Sources consisting of many / few words with frequency 1
6.4.2.4 Sources with low / high average word length of rare words