In the context of text analysis software, a query refers to a unit of analysis applied to the text data in a dataset. Each line in an Excel file containing text is considered one query. The character limit for each query is set at 1000 characters, and if a row exceeds this limit, it is treated as a separate query.
- The number of queries is influenced by the choice of classifiers and text analysis features.
- Selecting a sentiment classifier adds one query.
- Opting for text analysis on popular topics adds a second query.
- Choosing multi-label classification adds one more query.
- Selecting an automatic classifier introduces an additional query.
Therefore, the total number of queries expended is determined by the combination of chosen classifiers and features when uploading a dataset for analysis. This approach helps manage and assess the complexity and scope of text analysis tasks.
Now, let's see an example:
Here is a dataset that you can also download from Kimola's GitHub repository of NLP datasets. This dataset contains Amazon reviews of Apple's Magic Mouse and there are 271 rows on the file.
After uploading a dataset to the home screen of Kimola Cognitive and choosing classifiers, it shows how many queries will be spent. Now, let's upload this dataset and see how many queries it will spend:
- As you remember, we had 271 rows in the file.
- There must some reviews that exceed 1000 character threshold. So the file will spend 360 queries for each text analysis feature.
- You will be able to see this screen once you upload your dataset and follow the steps.
Visit Kimola Cognitive dashboard to create your report right away.
Was this article helpful?
Tell us about your thoughts and experiences regarding the article.