Differences between revisions 42 and 50 (spanning 8 versions)

Aqqu

Contents

Aqqu

Description

Question answering from Freebase as described in the CIKM 2015 publication. The code below also contains some improvements (neural network, performance) that came after the publication. The public code also contains a README that describes how to download, install, train and run the system. Below describes how to setup the demo (for which the code is not public).

Code

The main repository is public on GitHub:

https://github.com/ad-freiburg/aqqu

There is also an internal repository that should be kept in sync with the GitHub one but may contain branches for internal testing/conferences:

https://ad-git.informatik.uni-freiburg.de/ad/Aqqu

The Web UI is currently only available internally. It should probably made public too:

https://ad-git.informatik.uni-freiburg.de/ad/aqqu-webserver

Previously Elmar kept an internal Aqqu version on BitBucket this is very old and stale though:

https://bitbucket.org/elmar-haussmann/aqqu .

Similarly there is an old BitBucket repository for the Web UI:

https://bitbucket.org/elmar-haussmann/aqqu-webserver .

Instances

The recommended way to set up an Aqqu instances is described in the README on GitHub and uses nvidia-docker (or docker when not using a GPU).

The below instances were set up be me (Niklas) so they might need a sudo -u schnelle -i before to get the right permissions for the repositories.

This method is used on the following Aqqu instances:

On titan (http://titan.informatik.privat:8300)
- Using the QLever freebase-aqqu instance listed here:
- Under /local/ssd2/aqqu
On vulcano (http://vulcano.informatik.privat:8090):
- First trained on the sqtrain dataset then on wqsptrain and free917train
- Under /local/raid/ad/schnelle/aqqu
- Using the Virtuoso SPARQL Engine docker-compose setup
  - docker-compose up -d in /local/raid/ad/schnelle/virtuoso-compose
  - https://github.com/ad-freiburg/virtuoso-compose

These instances can be used with the following Aqqu Web UI instances

On titan (http://titan.informatik.privat:5454, http://titan.informatik.privat:5455 )
- Under /local/ssd2/aqqu_webui/
- Launched with a single docker-compose up -d --scale webui=2

Old Aqqu instance (via Elmar's home)

Start as follows on metropolis:

ssh metropolis
sudo su haussmae
cd /home/haussmae/demos/aqqu-demo
source venv/bin/activate   # was: activate aqqu
PYTHONPATH=$(pwd):$PYTHONPATH python webserver/translation_webserver.py

Virtuoso instance (using Docker Compose)

2018-04-02: Virtuoso instance for Aqqu runs at http://vulcano.informatik.privat:9000/sparql this is proxied to http://aqqu.informatik.uni-freiburg.de/sparql/

First make sure you have Docker + Docker Compose set up as described here: https://docs.docker.com/compose/install/ also make sure that your user is in the *docker group*

If you are setting up a new instance first check out the repository. If this is an existing instance just switch to its checked out virtuoso-compose (e.g. at vulcano:/local/raid/ad/schnelle/virtuoso-compose)

git clone https://ad-git.informatik.uni-freiburg.de/ad/virtuoso-compose.git
cd virtuoso-compose

Then start virtuoso with varnish as a caching proxy as follow (the resulting server listens on port 9000). Note that this is the same command whether you set up initially or manually launch the instance

docker-compose up -d

To view the log run

docker-compose logs -f

Virtuoso instance (via Elmars home)

Start virtuoso as follows. Server listens on port 8999. (Alternatively follow the instructions from scratch instructions below)

ssh metropolis
sudo su haussmae
cd /home/haussmae/keyword-translation
make start-virtuoso-mini-cai

Start the HTTP proxy as follows (kill old process manually, listens on port 9000):

ssh metropolis
sudo su haussmae
cd /home/haussmae/keyword-translation
make start-varnish

Parser

Aqqu uses a parser to get part-of-speech tags of query words. The parser is accessed via HTTP Api calls. To start the parser server:

ssh metropolis
sudo su haussmae
cd /home/haussmae/keyword-translation
make start-parser

The port is configured in the corenlp-frontent/build.xml. The API can be accessed like this: http://metropolis.informatik.uni-freiburg.de:4000/parse/?text=This%20is%20a%20test%20sentence.

Run the new Aqqu version (with NN)

Start on titan (requires GPU):

ssh titan
sudo su haussmae
cd /home/haussmae/aqqu-bitbucket
source activate aqqu
PYTHONPATH=$(pwd):$PYTHONPATH python webserver/translation_webserver.py

Runs on port 5454 on titan now. However, titan is not available from outside the uni network.

To start a port-forwarding from metropolis:

ssh metropolis
sudo su haussmae
cd /home/haussmae/temp/nc/python-port-forwardt
python2 port-forward.py

The service is now available on metropolis:5454

How to update (any) Virtuoso with custom data

Grant access rights via the ISQL tool as follows:

data/virtuoso/install/bin/isql localhost:1112 dba dba
grant execute on SPARQL_INSERT_DICT_CONTENT to SPARQL_UPDATE;
grant execute on SPARQL_INSERT_DICT_CONTENT to "SPARQL";
grant execute on SPARQL_DELETE_DICT_CONTENT to SPARQL_UPDATE;
grant execute on SPARQL_DELETE_DICT_CONTENT to "SPARQL";

Complex example SPARQL query from "Programmieren in C++, SS 2016, Ü10 (all action or animation movie with their release date, genre, director, production company, and rating):

PREFIX fb: <http://rdf.freebase.com/ns/>
 SELECT DISTINCT ?fn, ?y, ?gn, ?dn, ?pn, ?rn where {
 ?f fb:type.object.type fb:film.film .
 ?f fb:film.film.initial_release_date ?y .
 ?f fb:film.film.genre ?g .
 ?f fb:film.film.directed_by ?d .
 ?f fb:film.film.production_companies ?p .
 ?f fb:film.film.rating ?r .
 ?f fb:type.object.name ?fn .
 ?g fb:type.object.name ?gn .
 ?d fb:type.object.name ?dn .
 ?p fb:type.object.name ?pn .
 ?r fb:type.object.name ?rn
 FILTER(lang(?fn)='en')
 FILTER(lang(?gn)='en')
 FILTER(lang(?dn)='en')
 FILTER(lang(?pn)='en')
 FILTER(lang(?rn)='en')
 FILTER(?gn='Action Film'@en OR ?gn='Animation'@en)
}

Data

All of the required data to run Aqqu is part of the materials (see above). It is located in the data subfolder. The scripts to create this data are part of the (old) keyword-translation repository: https://bitbucket.org/onekonek/keyword-translation

-  ⇤ ← Revision 42 as of 2018-01-15 17:36:44 → 
  Size: 7396
  Editor: Niklas Schnelle
  Comment:
+   ← Revision 50 as of 2020-01-07 18:07:40 → ⇥
  Size: 6926
  Editor: Hannah Bast
  Comment:
-Deletions are marked like this.
+Additions are marked like this.
 Line 1:
-#acl Niklas Schnelle:read,write All:read
+#acl Niklas Schnelle:read,write Natalie Prange:read,write All:read
 Line 4:
+<<TableOfContents(3)>>
-Line 11:
+Line 13:
-Public !GitHub repository: https://github.com/elmar-haussmann/aqqu .
+The main repository is public on !GitHub:
-Line 13:
+Line 15:
-Internal git repository (contains work after publication, mainly neural net and performance improvements): https://bitbucket.org/elmar-haussmann/aqqu .
+https://github.com/ad-freiburg/aqqu
-Line 15:
+Line 17:
-Internal git repository for the web-UI (we didn't put that public): https://bitbucket.org/elmar-haussmann/aqqu-webserver .
+There is also an internal repository that should be kept in sync with the !GitHub one but may contain branches for internal testing/conferences:
-Line 17:
+Line 19:
-== Demo ==
+https://ad-git.informatik.uni-freiburg.de/ad/Aqqu
-Line 19:
+Line 21:
-=== Aqqu instance ===
+The Web UI is currently only available internally. It should probably made public too:
-Line 21:
+Line 23:
--06-30: runs under http://metropolis.informatik.uni-freiburg.de:5455
+https://ad-git.informatik.uni-freiburg.de/ad/aqqu-webserver

Previously Elmar kept an internal Aqqu version on !BitBucket this is very old and stale though: 

https://bitbucket.org/elmar-haussmann/aqqu .

Similarly there is an old !BitBucket repository for the Web UI:

https://bitbucket.org/elmar-haussmann/aqqu-webserver .

== Instances ==
The recommended way to set up an Aqqu instances is described in the [[https://github.com/ad-freiburg/aqqu/blob/master/README.md|README]] on !GitHub and uses `nvidia-docker` (or `docker` when not using a GPU).

The below instances were set up be me (Niklas) so they might need a `sudo -u schnelle -i` before to get the right permissions for the repositories.

This method is used on the following '''Aqqu''' instances:

 * On '''titan''' ([[http://titan.informatik.privat:8300]])
  * Using the QLever '''freebase-aqqu''' instance listed [[Projects/QLever#Instances|here]]:
  * Under `/local/ssd2/aqqu`
 * On '''vulcano''' ([[http://vulcano.informatik.privat:8090]]):
  * First trained on the `sqtrain` dataset then on `wqsptrain` and `free917train`
  * Under `/local/raid/ad/schnelle/aqqu`
  * Using the Virtuoso SPARQL Engine `docker-compose` setup
   * `docker-compose up -d` in `/local/raid/ad/schnelle/virtuoso-compose`
   * [[https://github.com/ad-freiburg/virtuoso-compose]]

These instances can be used with the following '''Aqqu Web UI''' instances

 * On '''titan''' ([[http://titan.informatik.privat:5454]], http://titan.informatik.privat:5455 )
  * Under `/local/ssd2/aqqu_webui/`
  * Launched with a single `docker-compose up -d --scale webui=2`

== Old Aqqu instance (via Elmar's home) ==
-Line 33:
+Line 68:
+=== Virtuoso instance (using Docker Compose) ===
2018-04-02: Virtuoso instance for Aqqu runs at http://vulcano.informatik.privat:9000/sparql 
this is proxied to http://aqqu.informatik.uni-freiburg.de/sparql/

First make sure you have Docker + Docker Compose set up as described here:
https://docs.docker.com/compose/install/
also make sure that your user is in the *docker group*

If you are setting up a new instance first check out the repository. If this is an existing instance just
switch to its checked out virtuoso-compose (e.g. at vulcano:/local/raid/ad/schnelle/virtuoso-compose)
{{{
git clone https://ad-git.informatik.uni-freiburg.de/ad/virtuoso-compose.git
cd virtuoso-compose
}}}

Then start virtuoso with varnish as a caching proxy as follow (the resulting server listens on port 9000).
Note that this is the same command whether you set up initially or manually launch the instance
{{{
docker-compose up -d
}}}

To view the log run
{{{
docker-compose logs -f
}}}
-Line 34:
+Line 94:
--06-30: Virtuoso instance for Aqqu runs unter http://metropolis.informatik.uni-freiburg.de:9000/sparql .
-Line 54:
+Line 112:
-}}}

=== Virtuoso instance (from scratch) ===

Start virtuoso as follows. Server listens on port 8999.

{{{
mkdir virtuoso-freebase
cd virtuoso-freebase
# get dependencies
sudo apt install bison flex gperf libssl-dev
# get the source
git clone https://github.com/openlink/virtuoso-opensource.git
cd virtuoso-opensource
git checkout stable/7
./autogen.sh
./configure --prefix=$(pwd)/virtuoso/install
}}}
check that everything ran without errors. Also make sure, that no virtuoso is already running
on this machine. The following steps intermediately start an instance and
fail otherwise
{{{
make -j && make install
cd ..
# get the data
wget "http://elba.informatik.uni-freiburg.de/freebase-qa/data/virtuoso.tar.gz" && tar xvfz virtuoso.tar.gz
# run in tmux
tmux
./virtuoso-opensource/virtuoso/install/bin/virtuoso-t -f +configfile virtuoso-db/virtuoso.ini
# ctrl+b+d
}}}

To get better performance through more aggressive caching we add varnish as a caching reverse proxy in front of virtuoso.

Install it from the repo to get security updates without hassle
{{{
sudo apt install varnish
}}}
For some reason Ubuntu thinks it should run this newly installed
service with a default config (it really isn't) so disable it again
{{{
sudo systemctl disable varnish.service # so it doesn't start again
sudo systemctl stop varnish.service # so the current instance is stopped
}}}
Now we need a config
{{{
mkdir varnish
cat << 'EOF' > varnish/varnish.vcl
#
# This puts varnish in front of virtuoso.
#
vcl 4.0;

# Default backend definition. Set this to point to your content server.
backend default {
    .host = "localhost";
    .port = "8999";
    .connect_timeout = 90s;
    .between_bytes_timeout = 30s;
}

sub vcl_recv {
    # Happens before we check if we have this in cache already.
    #
    # Typically you clean up the request here, removing cookies you don't need,
    # rewriting the request, etc.

    # Just put everything in cache.
    return(hash);
}

sub vcl_backend_response {
    # Happens after we have read the response headers from the backend.
    #
    # Here you clean the response headers, removing silly Set-Cookie headers
    # and other mistakes your backend does.

    # We set TTL to a looong time.
    set beresp.ttl = 99999h;

    if(beresp.status == 404) {
          # Cache 404 responses for 15 seconds
          set beresp.ttl = 15s;
          set beresp.grace = 15s;
    }
}

sub vcl_deliver {
    # Happens when we have all the pieces we need, and are about to send the
    # response to the client.
    #
    # You can do accounting or modifying the final object here.
    return(deliver);
}
EOF
}}}

And finally we start it. Don't forget to replace '''vulcano''' with your hostname
{{{
varnishd -a vulcano:9000 -f varnish/varnish.vcl -F -n /tmp -s malloc,10G -p http_resp_size=10000000 -p http_req_size=1000000 -p http_resp_hdr_len=1000000 -p http_req_hdr_len=1000000