Using Chatbots to Assist Communication in Collaborative Networks: 19th IFIP WG 5.5 Working Conference on Virtual Enterprises, PRO-VE 2018, Cardiff, UK, September 17-19, 2018, Proceedings
Novel technological possibilities enable a better communication and
knowledge exchange within collaborative networks. The paper indicates that
the potential of enterprise social network is by far not exploited yet. Although
systems are connected to each other, and the conditions for frictionless data
exchange are created, there is a lack of flexible possibilities to use the data
stock within the network efficiently and user-friendly. Chatbots provide a
possibility to meet this challenge. Nowadays, chatbots are used to improve
customer communication and simplify the daily routine in consumers’ lives.
Within collaborative networks, their use and benefits had not been fully
discovered yet. This paper examines current chatbot technologies and
implements a use case driven prototype to show the benefits of chatbots within
enterprise social networks for (internal) communication by smart combining
data across collaborative networks.
Similaire à Using Chatbots to Assist Communication in Collaborative Networks: 19th IFIP WG 5.5 Working Conference on Virtual Enterprises, PRO-VE 2018, Cardiff, UK, September 17-19, 2018, Proceedings
Similaire à Using Chatbots to Assist Communication in Collaborative Networks: 19th IFIP WG 5.5 Working Conference on Virtual Enterprises, PRO-VE 2018, Cardiff, UK, September 17-19, 2018, Proceedings (20)
Call Girls in Mayapuri Delhi 💯Call Us 🔝9953322196🔝 💯Escort.
Using Chatbots to Assist Communication in Collaborative Networks: 19th IFIP WG 5.5 Working Conference on Virtual Enterprises, PRO-VE 2018, Cardiff, UK, September 17-19, 2018, Proceedings
1. Using Chatbots to Assist Communication in
Collaborative Networks
Christian Frommert, Anna Häfner, Julia Friedrich and Christian Zinke
2. About
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
Working Group ET: Efficient Technology Integration
Datafrom to Solution
CC – Siwat V - https://www.iconfinder.com/icons/2377629/business_computer_devices_mobile_phone_responsive_technology_icon
Smart Processes & Services
Digital Platforms
Knowledge Integration/Engineering
3. (1) Give an overview of current chatbot engines and their states
(2) Design a solution for better interaction and knowledge exchange
(3) Implement a chatbot into use-cases to demonstrate the benefits
Goals
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
5. Social networks in enterprises
Web-based platform which
allow user to contribute
persistent objects to a
shared pool
Enterprise
Social Network
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
Leonardi, Huysman et al. 2013 – Enterprise Social Media
6. Social networks in enterprises
Collabor
ation
Communica
tion
Messenger
Task
Management
Document
Management
Enterprise
Social Network
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
7. NLP (Natural Language Processing)
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
8. Natural Language Processing
“is a theory-motivate rang of
computational techniques for
automatic analysis and
representation of human language”
NLP
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
Cambria, E., White, B.: Jumping NLP Curves. A Review of Natural Language
Processing Research [Review Article]. IEEE Comput. Intell. Mag. (2014). doi:
10.1109/MCI.2014.2307227
NLU – Natural Language
Understanding
9. Natural Language Processing
“is a theory-motivate rang of
computational techniques for
automatic analysis and
representation of human language”
NLP
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
NLU – Natural Language
Understanding
NLG – Natural Language
Generation
10. Natural Language Processing
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
NLU – Natural Language
Understanding
Classification of the input
(Intention)
Extraction of variables
(Entities)
Tools: : Spacy, MITIE,
Duckling
12. Chatbots
Is a text-based software system
that interact with human users
in their natural language and
make decisions based on
predefined rules
Chatbot
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
Shawar, B.A., Atwell, E.: Different Measurements Metrics to Evaluate a Chatbot System. In: Proceedings of the
Workshop on Bridging the Gap: Academic and Industrial Research in Dialog Technologies, pp. 89–96.
Association for Computational Linguistics, Stroudsburg, PA, USA (2007)
13. Chatbot
CC GPL v2 Iconfinder ApS: https://www.iconfinder.com/iconfinder
Input
Output
Analyse
Context
Action
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
Chatbots
14. In a nutshell
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
19. Overview
21 chatbots are evaluated
Deeper analysis on on-
premise and interoperable
Open Source Solutions
(10)
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
21. The daily struggle: arrange a meeting
inefficient
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
time-consuming
annoying
22. From chatbot to ESN – the architecture
Rasahub, https://github.com/DServSys/Rasahub
Rasahub Humhub (ESN)Rasa_Core
Rasa_NLU Google Calendar
Bot answer
SQL Insert
User Message
Data from Calendar
SQL Select
Bot answer
Socket
User Message
Data
Socket
Data from Calendar
API
User Message
Intents
Entities
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
23. Train the model – conversation flow
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
24. Train the model
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
## intent:inform_calendar
* please make an appointment on [Wednesday](day)
* please make an appointment on [Friday](day)
* find an appointment on [Thursday](day)
* find an appointment on [Monday](day)
* is there a free date on [Friday](day)
* is there a free date on [Monday](day)
* please make an appointment on [Tuesday](day) for
[10 minutes](duration:10)
* find appointment on [Thursday](day) for [10
minutes](duration:10)
## Generated Story -908337105314369877
* inform_calendar{"day": "heute", "time": "2018-01-
04T00:00:00.000Z"}
- action_extract_calendarinfo
- slot{"lastBotUtterance": "utter_ask_duration"}
- slot{"extractedDayFrom": "2018-01-
04T00:00:00.000Z"}
- slot{"timeValid": "True"}
* inform_calendar{"duration": 1.0}
- action_extract_calendarinfo
- slot{"lastBotUtterance": "utter_ask_participants"}
- slot{"extractedDuration": "60"}
- slot{"durationValid": "True"}
26. Prove of concept
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
Calendar Chat
Partner 2
Calendar Chat
Partner 1
Conversation with
the bot New Calendar Chat
Partner 2
Finale meeting
arrangement
27. Thank you
Christian Zinke
zinke@infai.org
sbdigital.infai.org
The German Federal Ministry of Education and Research have funded the work leading
to this publication under grant number 02L15A070 (Project: SB:Digital). The
authors thank for the funding
Dr. Christian Zinke - zinke@infai.org | sbdigital.infai.org
Notes de l'éditeur
Web‐based platforms that allow workers to (1) communicate messages with specific coworkers or broadcast messages to everyone in the organization; (2) explicitly indicate or implicitly reveal particular coworkers as communication partners; (3) post, edit, and sort text and files linked to themselves or others; and (4) view the messages, connections, text, and files communicated, posted, edited and sorted by anyone else in the organization at any time of their choosing