```
...
requests.exceptions.ConnectionError: HTTPConnectionPool(host='localhost', port=9001): Max retries exceeded with url: /?properties=%7B%22outputFormat%22%3A+%22json%22%2C+%22annotators%22%3A+%22tokenize%2Cpos%2Clemma%2Cssplit%2Cparse%22%2C+%22ssplit.eolonly%22%3A+%22true%22%2C+%22tokenize.whitespace%22%3A+%22false%22%7D (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x00000215465682B0>: Failed to establish a new connection: [WinError 10061] No connection could be made because the target machine actively refused it'))
```
The possible cause is:
- In some devices, CoreNLP local web services is slowly started, so it is not yet ready when directly utilise by NLIMED. You can wait for a minute then rerun your command or code.
- In Windows the JVM maximum memory space is between 1500 and 1600 MB, while CoreNLP is ideally run on 4GB heap memory space.
Solution:
You may also start the services manually on command line or terminal. Go to your CoreNLP folder and run:
```
java -Xmx4g -cp "*" edu.stanford.nlp.pipeline.StanfordCoreNLPServer -port 9001 -timeout 15000
```
Then check the services availability on your web browser through link: http://localhost:9001/
When using CoreNLP parser, you may find the following error message: