Skip to content
GitLab
Explore
Sign in
Primary navigation
Search or go to…
Project
Wiki Rasa
Manage
Activity
Members
Labels
Plan
Issues
Issue boards
Milestones
Iterations
Wiki
Requirements
Code
Merge requests
Repository
Branches
Commits
Tags
Repository graph
Compare revisions
Snippets
Locked files
Build
Pipelines
Jobs
Pipeline schedules
Test cases
Artifacts
Deploy
Releases
Package Registry
Model registry
Operate
Environments
Terraform modules
Monitor
Incidents
Service Desk
Analyze
Value stream analytics
Contributor analytics
CI/CD analytics
Repository analytics
Code review analytics
Issue analytics
Insights
Model experiments
Help
Help
Support
GitLab documentation
Compare GitLab plans
Community forum
Contribute to GitLab
Provide feedback
Keyboard shortcuts
?
Snippets
Groups
Projects
Show more breadcrumbs
tm-chatbot
Wiki Rasa
Commits
f5ce2319
Commit
f5ce2319
authored
6 years ago
by
Lucas Schons
Browse files
Options
Downloads
Plain Diff
Merge branch '20-uberarbeitung-slides' into 'master'
adds version1.1 of slides with better formatting. Closes
#20
See merge request
!18
parents
39ca4c52
4249583c
No related branches found
No related tags found
1 merge request
!18
adds version1.1 of slides with better formatting.
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
docs/2018-12-14-presentation.md
+16
-26
16 additions, 26 deletions
docs/2018-12-14-presentation.md
with
16 additions
and
26 deletions
docs/2018-12-14-presentation.md
+
16
−
26
View file @
f5ce2319
...
...
@@ -12,11 +12,7 @@
---
## Zielsetzung
---
### Rasa Chatbot
### Zielsetzung
RASA ist ein Framework für die Nutzung von Chatbots. Das System muss mit Facts befüllt
werden, um Antworten generieren zu können.
...
...
@@ -38,11 +34,11 @@ Dies soll anhand der Wikipedia-Einträge von Physikern erprobt werden.
---
## Daten
##
#
Daten
*
Knapp 1000 Wikipedia Artikel
*
Über Wikipedia API
*
HTML
*
HTML
-Formatierung
*
Ca. 550 Wörter/Artikel
*
Großteil um 400 Wörter
...
...
@@ -54,14 +50,6 @@ Dies soll anhand der Wikipedia-Einträge von Physikern erprobt werden.
## Funktionsweise Rasa-Bots
*
`nlu.md`
- _Intents_
*
`stories.md`
- _Gesprächsverlauf_
*
`domain.yml`
- _Entitäten, Actions, Slots_
*
`nlu_config.yml`
- _Pipeline_
*
`actions.py`
- _findBirthplace()_
---
```
rasa
├── actions.py - Custom Actions
...
...
@@ -101,10 +89,12 @@ class ActionSearchIsAlive(Action):
```
bash
"
is
Albert
Einstein
still
alive"
2018-12-12 14:39:38 DEBUG rasa_core.processor - Received user message
'is Albert Einstein still alive' with intent '{'name'
:
'
is_alive'
,
'
confidence'
:
0.9826370477676392}' and entities '[{'start'
:
3, 'end'
:
18, 'value'
:
'
Albert
Einstein'
,
'
entity'
:
'
Physicist'
,
'
confidence'
:
0.9928692094590056,
2018-12-12 14:39:38 DEBUG rasa_core.processor -
Received user message 'is Albert Einstein still alive'
with intent '{'name'
:
'
is_alive'
,
'
confidence'
:
0.9826370477676392}' and entities '[{'start'
:
3, 'end'
:
18, 'value'
:
'
Albert
Einstein'
,
'
entity'
:
'
Physicist'
,
'
confidence'
:
0.9928692094590056,
'extractor'
:
'
ner_crf'
}]
'
```
...
...
@@ -113,11 +103,11 @@ class ActionSearchIsAlive(Action):
## Antworten des Bots
```
bash
2018-12-12 18:54:23 DEBUG rasa_core.processor - Predicted next action
'action_search_is_alive'
with prob 1.00.
2018-12-12 18:54:23 DEBUG rasa_core.processor - Action
'action_search_is_alive'
2018-12-12 18:54:23 DEBUG rasa_core.processor -
Predicted next action
'action_search_is_alive'
with prob 1.00. 2018-12-12 18:54:23 DEBUG
rasa_core.processor - Action
'action_search_is_alive'
ended with events
'['
SlotSet
(
key: is_alive, value: False
)
']'
2018-12-12 18:54:23 DEBUG rasa_core.processor - Bot utterance
"The Life status of Albert Einstein is False."
```
...
...
@@ -128,8 +118,8 @@ ended with events '['SlotSet(key: is_alive, value: False)']'
*
Information Extraction
*
Name-Entity-Recognition für "einfache" Intents: Orte, Datum etc.
*
word2vec oder Bootstrapping Relationship Extractors
, um
die "schwierigeren"
Intents: Forschungsgebiet bzw. verheiratetM
IT
, kindV
ON
etc.
*
word2vec oder Bootstrapping Relationship Extractors
für
die "schwierigeren"
Intents:
'
Forschungsgebiet
'
bzw.
'
verheiratetM
it'
,
'
kindV
on'
etc.
*
RASA trainieren
*
APIs um die einzelnen Teile zusammen zubringen
*
APIs um die einzelnen Teile zusammen zu
bringen
This diff is collapsed.
Click to expand it.
Preview
0%
Loading
Try again
or
attach a new file
.
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Save comment
Cancel
Please
register
or
sign in
to comment