We provide ready-to-use software libraries for most important text analytics tasks.
Award Winning Technology
- Winner of SemEval-2016: SENTIMENT ANALYSIS on English tweets
- Winner of EvalItalia-2016: SENTIMENT ANALYSIS on Italian tweets
- Top Rank at PAN-2016: AGE and GENDER DETECTION on English tweets(rank 5)
|Named Entity Recognition(NER)||✓||✓|
|Topic Categorization (for business-related news)||✓|
We can easily apply our technology to any other language. Usually we just need a good set of training examples in the new language. If no such training data exists, we can help you to generate it.
Our software libraries are mostly written in Python. They are ready-to-use and come with a simple and straightforward API.
Our models are trained and optimized on different text types, including tweets, news articles and product reviews.
- How much does it cost?
We offer different types of licences, depending on the number of transactions per month. Contact us for a price quote and a free trial.
- I have a very special type of texts. Can you adapt the models to my domain?
You just give us a set of examples with your texts, and we can easily adapt our models to your domain.
If your data is highly sensitive, we can even train our models on-premise, such that your data does not leave your company.
- Do you offer SaaS or a web service?
At the moment, we only offer software libraries that you host on your own servers.
- How fast is one request?
This obviously depends on the model, but a typical request runs within 1 second or below.
Our libraries are scalable by design: If you need higher throughput, you can easily run several instances in parallel.
- I have a task that cannot be solved with the models above. What can I do?
The models above are for standard tasks, which occur in various applications such as media monitoring or customer support. However, there are also many applications that cannot be solved “of-the-shelf”.
For these cases, we can run a custom projects where we develop a customized solution for you.
Please contact us to discuss potential approaches to your problem.