{"id":42470,"date":"2018-03-08T22:41:44","date_gmt":"2018-03-08T21:41:44","guid":{"rendered":"https:\/\/www.clickworker.com\/?page_id=42470"},"modified":"2023-02-28T15:15:58","modified_gmt":"2023-02-28T14:15:58","slug":"training-data-for-chatbots","status":"publish","type":"page","link":"https:\/\/www.clickworker.com\/case-studies\/training-data-for-chatbots\/","title":{"rendered":"Training Data for Chatbots – Case Study"},"content":{"rendered":"\t\r\n\t
\r\n\t\t
\r\n\t\t\t
\r\n\t\t\t\t
\r\n\t\t\t\t\t\t

Training Data for Chatbots<\/h1>\r\n\t\t\t\t<\/div>\r\n\t\t\t<\/div>\r\n\t\t\t
\r\n\t\t\t\t
\"Training<\/div>\r\n\t\t\t\t
\r\n\t\t\t\t\t\r\n\t\t\t\t\t

Case Study \u2013 Creation of IT Support Questions in Text Format to Train an Artificial IT Service Desk Agent.<\/strong><\/p>\r\n\t\t\t\t\t

Thousands of Clickworkers formulate possible IT support inquiries based on given IT user problem cases. This creates a multitude of query formulations which demonstrate how real users could communicate via an IT support chat. With these text samples a chatbot can be optimized for deployment as an artificial IT service desk agent, and the recognition rate considerably increased.<\/p>\r\n\r\n\t\t\t\t\t <\/i> Get in touch with us!<\/a>\r\n\t\t\t\t\t\r\n\t\t\t\t\t<\/i> +1 (212) 878-6686<\/a>\r\n\t\t\t\t\t<\/i> +49 201 95971830<\/a>\r\n\t\t\t\t\t\r\n\r\n\t\t\t\t<\/div>\r\n\t\t\t<\/div>\r\n\t\t<\/div>\r\n\t<\/section>\r\n\t\r\n\t

<\/div>\r\n\r\n\t
\r\n\t
\r\n\t
\r\n\t
\r\n\t \"Training\r\n\t <\/div>\r\n\t
\r\n\t

The Challenge<\/h2>\r\n\t

As a rule chatbots access canned knowledge databases, in which answers to diverse questions are\r\n\t recorded. The principal challenge when programming chatbots is correctly recognizing the users\u2019\r\n\t questions, classifying them accurately in the database and issuing the correct answer, or asking\r\n\t valid follow-up questions if required. In addition, chatbots learn from each new inquiry. The more\r\n\t requests a chatbot has processed, the better trained it is. The knowledge database is continually\r\n\t expanded, and the bot\u2019s detection patterns are refined.<\/p>\r\n\r\n\t

\r\n\t Depending on the field of application for the chatbot, thousands of inquiries in a specific subject\r\n\t area can be required to make it ready for use. Moreover, a large number of additional queries are\r\n\t necessary to optimize the bot, working towards the goal of reaching a recognition rate approaching\r\n\t 100%.\r\n\t <\/p>\r\n\t <\/div>\r\n\r\n\t <\/div>\r\n\t

<\/div>\r\n\t\t\t\t\r\n\t\t\t
\r\n\t
\r\n\t

The Solution<\/h2>\r\n\t

Our Clickworkers have reformulated 500 existing IT support queries in seven languages,\r\n\t and so have created multiple new variations of how IT users could communicate with a support\r\n\t chatbot. Each predefined question is restated in three versions with different perspectives\r\n\t (neutral, he, she) for those languages that differentiate noun genders, or in two versions for\r\n\t languages that don\u2019t.<\/p>\r\n\t\t\t\t<\/div>\r\n\t\t

\r\n\t \"Training\r\n\t <\/div>\t\t\t\t\t\r\n\t\t\t<\/div>\r\n\t
\r\n\t
\r\n\t\t\t\t\t\t\r\n\t
\r\n\t
\r\n\t

Example:<\/p>\r\n\t \r\n\t \r\n\t \r\n\t \r\n\t \r\n\t \r\n\t \r\n\t
Preset sentence:<\/td>\r\n\t \u201cI can\u2019t log in.\u201d<\/td>\r\n\t <\/tr>\r\n\t
Neutral \u2013 restated:<\/td>\r\n\t \u201cA login is not possible.\u201d<\/td>\r\n\t <\/tr>\r\n\t
He\/She \u2013 restated:<\/td>\r\n\t \u201cMy colleague can\u2019t log in.\u201d<\/td>\r\n\t <\/tr>\r\n\t
<\/td>\r\n\t \u201cMy boss cannot log in.\u201d or<\/td>\r\n\t <\/tr>\r\n\t
<\/td>\r\n\t \u201cOur new employee can\u2019t log in.\u201d<\/td>\r\n\t <\/tr>\r\n\t <\/tbody>\r\n\t <\/table>\r\n\t

In an additional job type, Clickworkers formulate completely new queries for a fictitious IT\r\n\t support. For this task, Clickworkers receive a total of 50 different situations\/issues.<\/p>\r\n\r\n\t <\/div>\r\n\t <\/div>\r\n\r\n\t <\/div>\r\n\t\t\t\t\r\n\t\t\t\t

\r\n\t
\r\n\t
\r\n\t

Example:<\/p>\r\n\t \r\n\t \r\n\t \r\n\t \r\n\t
The user receives one of the following error messages<\/td>\r\n\t <\/tr>\r\n\t
1) \u201cNo user account exists.\u201d or<\/td>\r\n\t 2) \u201cNo user account found.\u201d etc.)<\/td>\r\n\t <\/tr>\r\n\t \r\n\t <\/tbody>\r\n\t <\/table>\r\n\t

\u2026 along with sample sentences which echo inquiries from company employees to an IT support unit. Per language and issue, 20 Clickworkers each formulate six questions in the first person, the way they would ask them in an IT support chat, to describe the given problem<\/p>\r\n\t\t\t\t\t\t\t

With the help of the numerous possible query formulations, the manufacturer trains the chatbot specifically for use as an IT service desk agent, and considerably increases the recognition rate and quality of the bot.<\/p>\r\n\t\t\t\t\t\t\t

More about our service \u201cTraining data for AI systems<\/a>\u201c<\/p>\r\n\r\n\t <\/div>\r\n\t <\/div>\t\t\t\t\r\n\t\t\t\t<\/div>\r\n\t <\/div>\r\n\t <\/div>\r\n\t<\/section>\r\n\r\n\r\n

\r\n\t
\r\n\t\r\n\t
\r\n\t\t
\r\n\t\t\t

Project Data<\/h2>\r\n\t\t<\/div>\r\n\t<\/div>\r\n\t
\r\n\t\t
\r\n\t\t\t
\r\n\t\t\t\t
\r\n\t\t\t\t\t
\r\n\t\t\t\t\t\t

Job \u201cReformulation\u201d<\/h3>\r\n\t\t\t\t\t\t