Hi, I am trying to follow the twitter clustering example from Mahout in Action. I have clustered users using the StandardAnalyzer, but I would like to see how the TwitterAnalyzer effects the results.
I can't figure out how to get it working though. I am stuck on the vectorizing process. I have a copy of mia-0.7.jar which does contain TwitterAnalyzer.class and I am trying to add it to the classpath before running the seq2sparse command. This is what my command looks like currently: sudo /opt/mahout/bin/mahout seq2sparse -s 5 -a mia.clustering.ch12.TwitterAnalyzer -chunk 800 -o /output_directory/vectors_directory -i /input_directory/twitter_seqfiles/part-r-00000 -md 3 -x 90 -wt tfidf -ml 10 -ng 3 -seq -nv I have tried adding "java -cp /home/efx/Desktop/mia-0.7.jar:" to the beginning of the command but I'm not having much luck. Apologies if this question is basic to you (it isn't to me). Many thanks, Will P.S. (I know mahout 0.8 is now available but I've been working with 0.7 until now and my project is nearly over)
