[Clustering] Solution training fails, as the Input fields are not really natural text, such as [First name], [Last name], [Street], [City], etc and the associated word corpus is based on natural text fields, such as [short_description] and [description]SummaryWhen training a Clustering solution that uses Input fields such as [First name], [Last name], [Street], [City], etc, which are not really natural text fields, it will result in an "Error while training solution", when it is associated with a natural text-based word corpus - Solution Training Failed. Ask Support to use log key XXXXXX to investigate trainer logs further. The log key XXXXXX is a number that helps Technical Support identify the error in the Training Server logs, which can only be accessed by Technical Support and it will show the following exception - INFO | jvm 8 | 2020/11/19 02:32:00.837 | Failed to train-model for Model-Parameters: [null, KMEANS, {1}, Geometric Mean Accuracy] due to: There is no NeuralEmbeddingModel associated with file: embedding_wv_model_WordVectors.zipINFO | jvm 8 | 2020/11/19 02:32:00.837 | SEVERE: DxC_ML: tid=47, msg=SE0060:Solution Training Failed. Ask Support to use log key 181111 to investigate trainer logs further.InstructionsThe error is caused by using a word corpus for a Clustering solution that uses natural text fields, such as [short_description] and [description] on table [incident], but the Clustering solution contains Input fields that are not really natural text, such as [First name], [Last name], [Email], [Home phone], [Street], [City], [Zip / Postal code], [State / Province] and [Country]. Therefore, the word corpus should be the same as the Input fields on the Clustering solution to train this type of Clustering solution successfully. However, for these types of Clustering solutions, we recommend using the the Levenshtein Distance method based clustering using DBScan, as this will work better for these types of Clustering solutions, where you are trying to cluster on fields that are not natural text, as the Levenshtein Distance method works better than the Word embedding algorithm Word2Vec. If you train your solution using the Levenshtein Distance method, you don't need to use a word corpus in your clustering solution. Please review our documentation on Configure Connect Component algorithm and Levenshtein Distance method for a clustering solution [Paris] for further information on the different algorithms that can be used for your Clustering solution training to optimize the trained solution, based on the type of Input fields that have been used.