Service and Support
Weka's customer service is available 24/7. Some find technical support lacking, with delays noted. Forums and groups provide community-driven help, rated highly by many. There are mixed opinions, with some praising responsiveness and interaction, others find support insufficient, rating it five to six. Weka's users benefit from communal resources and documentation rather than direct support channels. Some rely on platforms like Stack Overflow for additional help beyond official support.
Deployment
Many find Weka's initial setup easy and intuitive, requiring little time, often under 15 minutes. The installation process is straightforward, especially for non-programmers, with effective documentation and tutorials. Data size impacts implementation time, and Weka handles small to medium datasets efficiently. Some note challenges with large datasets and initial data loading, but these become manageable with familiarity. Users praise the cloud deployment and value Weka's simplicity over alternatives like R.
Scalability
Weka's scalability is mixed. Some users report challenges with large datasets and note limitations in handling multiple users without horizontal scaling. In contrast, others praise its vertical scalability and stable performance once deployed. Users who work alone or with smaller data find it sufficient, while those requiring integration express concerns about scalability. It is effective for single machine learning tasks and praised for ease of use, but concerns remain regarding handling large datasets efficiently.
Stability
Weka is largely considered stable with high ratings, often described as reliable without significant issues or crashes. Stability is sometimes influenced by the selection of predictor variables, impacting classification accuracy. Users mention the need for better system resources for optimal performance. Occasionally, crashes are noted, but many find Weka satisfactory for their needs. Improvements in dynamic output and documentation are suggested to enhance functionality further.