Processing service notes
Example techniques for processing service notes are described. In an example, labeled service notes, associated with fuser units of a plurality of image rendering devices, are processed to generate a vector corresponding to each of the labeled service notes, a labeled service note comprising natural language text describing an error event and a corresponding service activity associated with a fuser unit, wherein the labeled service note is assigned a label based on a category of failure of the fuser unit. Based on the processing, a relationship between vectors and labels corresponding to the labeled service notes is generated.
1. A method comprising:
obtaining labeled service notes associated with fuser units of a plurality of image rendering devices, each labeled service note comprising natural language text describing an error event and a corresponding service activity associated with a fuser unit of an image rendering device, wherein each labeled service note is assigned a label based on a category of failure of the fuser unit;
processing the labeled service notes to generate vectors respectively corresponding to the labeled service notes;
generating a function mapping a relationship between the vectors and the labels of the labeled service notes; and
determining a label for an unlabeled service note based on the function.
2. The method as claimed in claim 1 , wherein processing the labeled service notes comprises, for each labeled service note:
analyzing the labeled service note to determine if an error code relating to the error event described in the labeled service note is included in the labeled service note; and
extracting the error code based on the analyzing to generate the vector corresponding to the labeled service note.
3. The method as claimed in claim 1 further comprising:
retrieving service notes associated with the plurality of image rendering devices, from a database, wherein the database stores details of error events and service activities relating to the plurality of image rendering devices; and
filtering the service notes associated with the fuser units of the plurality of image rendering devices.
4. The method as claimed in claim 3 further comprising:
capturing, in the database, the error events based on user inputs pertaining to the error events; and
capturing, in the database, the service activities based on remarks of technicians servicing the plurality of image rendering devices further to the error events.
5. The method as claimed in claim 3 , further comprising labeling, based on predefined rules, the filtered service notes associated with the fuser units to obtain the labeled service notes.
6. The method as claimed in claim 1 , further comprising translating the unlabeled service note into English language if the unlabeled service note is in a language other than English.
7. A machine learning device comprising:
a natural language processing engine to:
process labeled service notes associated with fuser units of a plurality of image rendering devices to generate vectors respectively corresponding to the labeled service notes, each labeled service note comprising natural language text describing an error event and a corresponding service activity associated with a fuser unit, wherein each labeled service note is assigned a label based on a category of failure of the fuser unit; and
a learning engine to:
learn a relationship between the vectors and the labels of the labeled service notes.
8. The machine learning device as claimed in claim 7 , further comprising a label prediction engine to determine a label for an unlabeled service note based on the relationship learned by the learning engine.
9. The machine learning device as claimed in claim 7 , further comprising, a fuser event prediction engine trained using service notes labeled based on the relationship between the vectors and the labels, to predict malfunctioning event for a given fuser unit.
10. The machine learning device as claimed in claim 7 , further comprising a communication engine to:
retrieve, from a database storing details of error events and service activities related to the plurality of image rendering devices,
inputs of users pertaining to the error events; and
remarks of technicians servicing the plurality of image rendering devices upon occurrence of the error events; and
combine the inputs of users pertaining to the error events with the remarks of technicians to obtain service notes.
11. The machine learning device as claimed in claim 10 , wherein the communication engine is to filter the service notes associated with the fuser units of the plurality of image rendering devices.
12. The machine learning device as claimed in claim 10 , further comprising a label input engine to assign labels to the service notes associated with the fuser units to obtain the labeled service notes based on user inputs.
13. The machine learning device as claimed in claim 8 , further comprising a translation engine to:
determine the unlabeled service note to be in a language other than English; and
translate the unlabeled service note into English.
14. A non-transitory computer-readable medium comprising instructions executable by a processing resource to:
obtain a service note associated with a fuser unit of an image rendering device, the service note comprising natural language text describing an error event and a corresponding service activity relating to the fuser unit; and
determine a label for the service note, wherein the label is indicative of a category of failure of the fuser unit, wherein
the label is generated based on a function mapping a relationship between labels and corresponding labeled service notes associated with fuser units of a plurality of image rendering devices.
15. The non-transitory computer-readable medium as claimed in claim 14 , further comprising instructions executable by the processing resource to translate the obtained service note into English.