Imagine a workbench
You set up your data set and one by one take the tools to bring it into the desired state. This is master data harmony. The data is uploaded and gradually transformed into the desired state. We make the right selection with you and demonstrate it in the trial cleansing, at no cost to you.
Our toolbox
Data Quality Assessment
Every project begins with the data quality assessment. How complete is the data set? How homogeneous is the formatting of the attributes? This and much more can be determined quickly using the data quality assessment so that there is transparency for the cleansing process.
Data Quality Assessment
Every project begins with the data quality assessment. How complete is the data set? How homogeneous is the formatting of the attributes? This and much more can be determined quickly using the data quality assessment so that there is transparency for the cleansing process.
Form reference designation
You are looking for the cover for your compaction units. But the result is only entries from the last 4 months - What were the covers called before? How many covers do you have in the system? Each department has done the naming independently, even after the introduction of central data management. You are looking at data from the last 14 years and need a standardized search term.
The basic designation is just that: an attribute that describes the core of the matter. "High-voltage cable", "adapter cable" and "ADAPT.-CAB." become "cable". The right level of granularity is defined individually for each project.
Form reference designation
You are looking for the cover for your compaction units. But the result is only entries from the last 4 months - What were the covers called before? How many covers do you have in the system? Each department has done the naming independently, even after the introduction of central data management. You are looking at data from the last 14 years and need a standardized search term.
The basic designation is just that: an attribute that describes the core of the matter. "High-voltage cable", "adapter cable" and "ADAPT.-CAB." become "cable". The right level of granularity is defined individually for each project.
Extract attributes
You are looking for the black plastic cover for your compaction units from 400 l/min upwards, but with all the covers it is almost faster to put them on again. The attribute you need is not maintained. In one of the 4 designation fields is the note, but each person has noted it differently. Your filters are not helping you. Not yet.
Attribute extraction retrieves attributes from existing texts and formats all entries according to a scheme of your choice, including existing entries. Finally you can find what you are looking for.
Extract attributes
You are looking for the black plastic cover for your compaction units from 400 l/min upwards, but with all the covers it is almost faster to put them on again. The attribute you need is not maintained. In one of the 4 designation fields is the note, but each person has noted it differently. Your filters are not helping you. Not yet.
Attribute extraction retrieves attributes from existing texts and formats all entries according to a scheme of your choice, including existing entries. Finally you can find what you are looking for.
Harmonize spelling (under development)
With the new basic designation, you have quickly found your cover. You will notice that there are some spelling variants in the designation texts, including the entry "Cover 400 L/min BLACK 72/5A". You recognize the responsible department and the production standard location in "72/5A" - good thing the designation field has not been overwritten. Yet you need a clean designation for your webstore. Correcting all 142,569 entries manually is out of the question, there is no capacity for that.
The spell checker is here to help. You find all spelling variants of cover and define your reference spelling. You exclude two variants that were recognized as similar from the change - these are other components. All 142,569 entries are then updated according to your specifications. Your webstore can go online.
Harmonize spelling (under development)
With the new basic designation, you have quickly found your cover. You will notice that there are some spelling variants in the designation texts, including the entry "Cover 400 L/min BLACK 72/5A". You recognize the responsible department and the production standard location in "72/5A" - good thing the designation field has not been overwritten. Yet you need a clean designation for your webstore. Correcting all 142,569 entries manually is out of the question, there is no capacity for that.
The spell checker is here to help. You find all spelling variants of cover and define your reference spelling. You exclude two variants that were recognized as similar from the change - these are other components. All 142,569 entries are then updated according to your specifications. Your webstore can go online.
Classify parts
Kürzlich haben Sie eine Architekturklasse eingeführt: Basierend auf den durch Master-Data-Harmony gefüllten Sachmerkmalsleisten tragen Sie Bauteil für Bauteil die 8-stellige Klasse ein, die Sie sich überlegt haben. Bald merken Sie, dass der Ansatz sehr lange dauern wird.
Hier kommt die Bauteilklassifikation zum Einsatz, die individuell mit Ihren Daten antrainiert wird. Anhand von händisch gepflegten Referenzdaten lernt der Code Ihre Anforderungen kennen und kann diese dann auf alle weiteren Daten anwenden. Im Handumdrehen haben Sie die neue Architekturklasse ausgerollt, ohne das Tagesgeschäft zu beeinträchtigen.
Classify parts
Kürzlich haben Sie eine Architekturklasse eingeführt: Basierend auf den durch Master-Data-Harmony gefüllten Sachmerkmalsleisten tragen Sie Bauteil für Bauteil die 8-stellige Klasse ein, die Sie sich überlegt haben. Bald merken Sie, dass der Ansatz sehr lange dauern wird.
Hier kommt die Bauteilklassifikation zum Einsatz, die individuell mit Ihren Daten antrainiert wird. Anhand von händisch gepflegten Referenzdaten lernt der Code Ihre Anforderungen kennen und kann diese dann auf alle weiteren Daten anwenden. Im Handumdrehen haben Sie die neue Architekturklasse ausgerollt, ohne das Tagesgeschäft zu beeinträchtigen.
Error detection
You have drawn the consequences of incomplete data maintenance: no component can be created without weight information ... After a few months you receive complaints, that weight information is no longer reliable. The Data Quality Assessment of Master-Data-Harmony shows you: 80% of new components weigh either 1 gram or 9 grams. You have placeholder entries. How are you supposed to find out, which entries are placeholders among all the components? You don't have the time to check each CAD model manually.
The error finder is here to help. You provide Master-Data-Harmony with a small data set for training, the quality of which you have verified. The error finder applies the training knowledge to all your data and highlights where the weight appears unusual. The task has come down to a managable level. A short time later, weights can be used again in your processes without hesitation.
Error detection
You have drawn the consequences of incomplete data maintenance: no component can be created without weight information ... After a few months you receive complaints, that weight information is no longer reliable. The Data Quality Assessment of Master-Data-Harmony shows you: 80% of new components weigh either 1 gram or 9 grams. You have placeholder entries. How are you supposed to find out, which entries are placeholders among all the components? You don't have the time to check each CAD model manually.
The error finder is here to help. You provide Master-Data-Harmony with a small data set for training, the quality of which you have verified. The error finder applies the training knowledge to all your data and highlights where the weight appears unusual. The task has come down to a managable level. A short time later, weights can be used again in your processes without hesitation.
Your data cleansing project
Step: 1
Trial cleansing
How good is your master data? Which steps are suitable for cleansing? In the preliminary coordination phase, we discuss a suitable cleansing process with you. You can see the added value for yourself with a free trial cleansing.
Step: 2
Data cleansing
When you place your order, we clean up your data using the defined tools. As required, it is possible to refine the data, e.g. if you want to adjust the formatting of the extracted attributes at a later date.
Step: 3
Reimporting
Once the data has been cleansed to your satisfaction, we provide you with the data so that it can be imported back into your system via bulk import.
Maintaining high data quality
In addition to initial data cleansing, you can also choose iterative updates. Your new data is brought up to the high quality standard you have become accustomed to through scaled data cleansing at the intervals of your choice. Contact us!
Further information
Master Data Harmony
... is your platform when it comes to the scaling cleansing of technical master data. The platform is hosted on German servers and used for order-based data cleansing. Through close contact with RWTH Aachen University, we ensure the use of modern methods for your data. We are constantly expanding our portfolio of cleansing tools.