Procsedings of the XVI EURALEX Intemational Congress: The User in Focus 'The exact number of the CCS corpus tokens is not known, but is should be somewhere between 1400 .000 and 2 100 000. The excerpts are of various lengths and multiplied so that they can be organized within two card-files (an example of excerpt cards can be seen in the Figure 1). AOATH : A 10 šion/6-4# Zaue — Sdiova ji B8 AONO 3 4 4 Mwmmá'- BeroAElić & diE H G kire Teudr" ŠA AUBANO < H aGTH ČACOH? BEMEM? K Hemugen? diziom. HM-a0c 68 AdNO B9BAH ddijio a ak adu k a'm ww%“ angelić, quubud Lattom £ račene a Wa , Figure 1: An excerpt card (used in the sources card-file and in the azbuka card-file). Along the excerpt cards run the so-called parallel cards with the variations of the lexical constituents of the excerpt as found in other sources (an example ofa parallel card is given in Figure2.). bO\'\' “ 340 stow a. 68 NONO 346j m iomwo S Aaty tžnje : NONO BM n w — > o Figure 2: A parallel card (used in the sources card-file and in the azbuka card-file). 'The first card-file is established according to the sources (informally,the sources card-file) and it con- tains approximately 420 000 cards. The second card-file is established according to the azbuka se- quence of the lemmas (informally,the azbuka card-file), with more than 400 000 cards.The azbuka 1224