diff --git a/services/data-hub/README.md b/services/data-hub/README.md
index 3dd233963..3d8112134 100644
--- a/services/data-hub/README.md
+++ b/services/data-hub/README.md
@@ -42,6 +42,10 @@ VERSION=latest npm run build:docker
 
 Kubernetes descriptors can be found in the [k8s](./k8s) directory.
 
+## Import & Merge
+
+Find more details [here](./src/minhash-poc/README.md) 
+
 ## Environment variables
 
 ### General
diff --git a/services/data-hub/src/minhash-poc/README.md b/services/data-hub/src/minhash-poc/README.md
index 142ecda8e..0b0e65f76 100644
--- a/services/data-hub/src/minhash-poc/README.md
+++ b/services/data-hub/src/minhash-poc/README.md
@@ -33,6 +33,6 @@ Then different data sets were checked for the existence of a common identity. Th
 
 The "minhashing" of small data sets basically works here and is suitable for establishing the same identities in different data sets.
 
-In contrast to the conventional application of these algorithms, the OIH's context tends to handle tiny-sized data sets. Each set usually contains a small amount of characters and therefore results more often in a meager set of shingles. In the end this can lead to accumulated false positives.
+In contrast to the conventional application of these algorithms, the OIH's context tends to handle tiny-sized data sets. Each set usually contains a small number of characters and therefore results more often in a meager set of shingles. In the end, this can lead to a higher amount of false positives.
 
 One possible conclusion is that automatic detection using the Minhash method should not be accepted without a further verification process. This should be done by algorithmic post-processing or plain human interaction on a suited merge interface.
\ No newline at end of file