Updated README file (3)

Author: sphl

README.md

@@ -91,10 +91,12 @@ to install and load all required R packages.
 
 #### Fuzzing Saturation / Stopping Criteria Analysis
 
-- `fuzzer_cov_data.csv`: ...
-- `fuzzer_cov_saturation.csv`: ...
-- `fuzzer_crashes.csv`: ...
+- `fuzzer_cov_data.csv`: Contains the code and bug coverage of various fuzzers at different time points (15-minute intervals) within 24-hour fuzzing campaigns.
+- `fuzzer_cov_saturation.csv`: Contains the times at which the fuzzing campaigns are terminated according to the different saturation-based stopping criteria examined in this study.
+- `fuzzer_crashes.csv`: Contains the fuzzer crashes, including the time points when they were triggered and the deduplication (column `Bug_ID`[^1]) for extracting the unique bugs.
 
 ### Pre-trained Vulnerability Prediction Models
 
 The pre-trained models can be found in the `./models` directory. Please refer to section `Green Fuzzing ... > ML-based Vulnerability Prediction > Vulnerability Prediction` in the `analysis.ipynb` notebook to see how these models can be used to predict potentially vulnerable functions.
+
+[^1]: Crashes with the same bug ID have the same top $N=3$ stack frames and are therefore considered to have the same underlying vulnerability.

analysis.ipynb

@@ -3219,7 +3219,7 @@
     "# corresponding line below (spoiler: the ratio of missed bugs hardly changes).\n",
     "fuzzer_crashes = read.csv(\"./data/fuzzer_crashes.csv\") |>\n",
     "    # filter(Seed_Type == \"non-empty\") |>\n",
-    "    distinct(Subject, Crash_ID) |>\n",
+    "    distinct(Subject, Bug_ID) |>\n",
     "    group_by(Subject) |>\n",
     "    summarise(N_Unique_Bugs = n())\n",
     "\n",

data/fuzzer_crashes.csv

@@ -1,4 +1,4 @@
-"Subject","Fuzzer","Seed_Type","Trial_Nr","File","Function","Line","Stack_Trace","Trace_Length","Trace_File","Timestamp","Crash_ID"
+"Subject","Fuzzer","Seed_Type","Trial_Nr","File","Function","Line","Stack_Trace","Trace_Length","Trace_File","Timestamp","Bug_ID"
 "JasPer","AFL","empty",2,"jas_malloc.c","jas_free",111,"jas_malloc.c:jas_free:111-jpc_dec.c:jpc_dec_cp_destroy:1557-jpc_dec.c:jpc_dec_destroy:1894-jpc_dec.c:jpc_decode:285-jp2_dec.c:jp2_decode:215-jas_image.c:jas_image_decode:372-jasper_fuzz.c:main:249",7,"%SRCROOT%/crash_traces/afl/jasper/seed_01/trial_02/crash_input_1633436959.trace","2021-10-05 14:29:19",61
 "JasPer","AFL","empty",2,"jas_seq.c","jas_seq2d_create",89,"jas_seq.c:jas_seq2d_create:89-jpc_dec.c:jpc_dec_tileinit:699-jpc_dec.c:jpc_dec_process_sod:556-jpc_dec.c:jpc_dec_decode:390-jpc_dec.c:jpc_decode:254-jp2_dec.c:jp2_decode:215-jas_image.c:jas_image_decode:372-jasper_fuzz.c:main:249",8,"%SRCROOT%/crash_traces/afl/jasper/seed_01/trial_02/crash_input_1633433709.trace","2021-10-05 13:35:09",110
 "JasPer","AFL","empty",2,"jpc_bs.c","jpc_bitstream_getbits",198,"jpc_bs.c:jpc_bitstream_getbits:198-jpc_t2dec.c:jpc_dec_decodepkt:314-jpc_t2dec.c:jpc_dec_decodepkts:454-jpc_dec.c:jpc_dec_process_sod:590-jpc_dec.c:jpc_dec_decode:390-jpc_dec.c:jpc_decode:254-jp2_dec.c:jp2_decode:215-jas_image.c:jas_image_decode:372-jasper_fuzz.c:main:249",9,"%SRCROOT%/crash_traces/afl/jasper/seed_01/trial_02/crash_input_1633436762.trace","2021-10-05 14:26:02",10