Add pagereactor target

Makefile

@@ -89,6 +89,7 @@ unsupervised2 := $(outdir)/unsupervised2.txt
 semi_supervised0 := $(outdir)/semi_supervised0.txt
 semi_supervised1 := $(outdir)/semi_supervised1.txt
 semi_supervised2 := $(outdir)/semi_supervised2.txt
+pagereactor0 := $(outdir)/pagereactor0.txt
 
 # ------------------------------------------------------------------------------
 
@@ -101,9 +102,9 @@ EM_MAIN := "try, All_BIC_ExplEM_Main; catch, end, exit"
 
 # ==============================================================================
 
-# perform baseline and exploratory clustering
+# perform baseline, exploratory and pagereactor clustering
 .DELETE_ON_ERROR:
-all: baseline explore
+all: baseline explore pagereactor
 
 # ==============================================================================
 
@@ -346,6 +347,16 @@ $(semi_supervised2): $(rid_fid_weight) $(rid_lid_score) $(qid_rid) \
 	$(PYTHON) $(srcdir)/exploratory.py $(FORMATTING_FLAGS) \
 		$(assgn_suffix) $(qid_rid) $(qid_eid) > $@
 
+# ------------------------------------------------------------------------------
+
+# pagereactor clustering
+.PHONY: pagereactor
+pagereactor: $(pagereactor0)
+
+# pagereactor output grouped by string
+$(pagereactor0): $(qid_tacid) | $(outdir)
+	cp $(qid_tacid) $@
+
 # ==============================================================================
 
 # create virtualenv

README

@@ -190,15 +190,30 @@ following exploreEM versions:
 Detailed information regarding the clustering algorithm and its parameters can 
 be found in the ExploreEM documentation.
 
-- "make" or "make all" is shorthand for calling "make raw", "make baseline" and
-"make explore".
+- "make pagereactor" performs clustering based on the pagereactor output. At the 
+moment, there is only one version:
+    -- pagereactor0:    String only. The assignment is performed as a two step
+                        process. In a first step, pagereactor entities whose 
+                        tacid is NULL and whose generic type is not NULL and not 
+                        OTHER are grouped by their wp14 name and assigned a nid.
+                        In a second step, those pagereactor entitites that have 
+                        either an eid or a nid are matched with their qid. The
+                        matching is performed based on string identity and
+                        appearance, e.g. if there are three tac queries with the 
+                        string "abraham_lincoln" and five pagereactor queries 
+                        with the string "abraham_lincoln" then the first three 
+                        of those pagereactor queries will be matched with the 
+                        qid's from the three tac queries.
+
+- "make" or "make all" is shorthand for calling "make raw", "make baseline", 
+"make explore", and "make pagereactor".
 
 Cleaning up
 -----------
 
 - "make clean" removes all the input generated by "make raw" (i.e., the data 
-directory) and the output generated by "make baseline" and "make explore" (i.e.,
-the output directory).
+directory) and the output generated by "make baseline", "make explore", and
+"make pagereactor" (i.e., the output directory).
 
 - "make cleandist" removes the same targets as "make clean" and, in addition
 to that, also removes the virtual environment. Note that in general "make clean"