correct readme

Author: huanchenz

README.md

@@ -1,7 +1,7 @@
 # RocksDB Equipped With SuRF
 
 ## Install Dependencies
-    sudo apt-get install build-essential cmake libsnappy
+    sudo apt-get install build-essential cmake libgtest.dev libsnappy-dev
     cd /usr/src/gtest
     sudo cmake CMakeLists.txt
     sudo make
@@ -15,52 +15,39 @@
     cmake -DWITH_SNAPPY=ON ..
     make -j 8
 
-## Generate Workload and RocksDB Instances
-The experiments presented in our
+## Generate Workloads and RocksDB Instances
+The experiment results presented in our
 [SIGMOD paper](http://www.cs.cmu.edu/~huanche1/publications/surf_paper.pdf)
-uses 100GB datasets. Because it takes a long time to run, we scale the experiments
-down to 2GB datasets (the 100GB experiment configs are still included in
-filter_experiment/filter_experiment.cc, commented out).
+uses 100GB datasets. Because it takes a long time to run, we scale
+down to 2GB datasets here (the 100GB configs are still included in
+**filter_experiment/filter_experiment.cc**, but commented out).
 
     cd filter_experiment
     python poisson.py
     mkdir data_no_filter data_bloom data_surf
 
-Run the executable "../build/filter_experiment/filter_experiment" will show the usage information:
+Run the executable **../build/filter_experiment/filter_experiment** will show the usage information:
     Usage:
     arg 1: path to datafiles
-    arg 2: filter type
-    0: no filter
-    1: Bloom filter
-    2: SuRF
-    3: SuRF Hash
-    4: SuRF Real
-    arg 3: compression?
-    0: no compression
-    1: Snappy
-    arg 4: use direct I/O?
-    0: no
-    1: yes
-    arg 5: query type
-    0: init
-    1: point query
-    2: open range query
-    3: closed range query
+    arg 2: filter type  0: no filter  1: Bloom filter  2: SuRF  3: SuRF Hash  4: SuRF Real
+    arg 3: compression? 0: no compression  1: Snappy
+    arg 4: use direct I/O?  0: no  1: yes
+    arg 5: query type  0: init  1: point query  2: open range query  3: closed range query
     arg 6: range size
     arg 7: warmup # of queries
 
-To initialize the RocksDB instances with no filter, bloom filters and SuRF (with real suffixes):
+To initialize the RocksDB instances with no filter, bloom filters and SuRFs (with real suffixes):
 
     ../build/filter_experiment/filter_experiment data_no_filter 0 1 0 0 0 0
     ../build/filter_experiment/filter_experiment data_bloom     1 1 0 0 0 0
     ../build/filter_experiment/filter_experiment data_surf      4 1 0 0 0 0
 
-## Run Benchmark
+## Run Benchmarks
 You may want to clear system cache (echo 3 | sudo tee /proc/sys/vm/drop_caches)
 before running each experiment. The source file for the experiments is
-"filter_experiment/filter_experiment.cc". To get correct I/O counts, you need
-to specify the device in function "getIOCount()" and "printIO()".
-filter_experiment.cc includes more configurations (e.g., specify filter sizes)
+**filter_experiment/filter_experiment.cc**. To get correct I/O counts, you need
+to specify the device in function **getIOCount()** and **printIO()**.
+**filter_experiment.cc** includes more configurations (e.g., specify filter sizes)
 to run different experiments besides the following examples:
 
     // point queries