Text analytics web services
Text analytics. Whether you are building an online marketplace, a social network app or a help desk, Textgain gives you insight into how people feel about your product or brand, based on their personal writing style. For example, ‘I love it!’ expresses a positive opinion, most likely written by a woman. Statistically, women use ‘I’ and ‘love’ more often than men.
Web services. Simply send text to our secure servers and get the analysis report in real-time, formatted as JSON. No need to install software or pay for upgrades. And we support many languages.
| Q |
https://api.textgain.com/en/age?q=I+love+it! |
|---|---|
| A | {"age": "25-", "confidence": 0.75} |
-
Language and genre identification
Identify the language and genre of a text.
-
Morpho-syntactic Analysis
Hyphenate words and count syllables. Identify sentence breaks and word types: nouns, verbs, adjectives, ...
-
Concept extraction
Identify keywords and named entities (proper names).
-
Sentiment analysis
Predict whether a text is objective or subjective, positive or negative. Fine‑grained analysis identifies product aspects: price, battery life, ...
-
Age & gender
Predict whether a text is written by an adolescent (<25) or an adult (>25), by a man or a woman.
-
Personality & education
Predict whether a text is written by an extraverted or introverted person, with basic or advanced writing skills.
Social media text profiling
Social media.
Right now, thousands of new tweets, reviews and articles are being written.
This is how people share their opinions and recommendations these days.
The amount of information is so vast
Text profiling. Textgain tells you who your users are (age, gender, education, personality, ...) and what they think, with an accuracy that rivals and surpasses humans, without asking users for private information. The AI is never stressed, bored or biased.
The figures show a live analysis (updated weekly) of tweets that mention Atari 800 or Commodore 64, an ongoing competition since the 1980s.
Case study: cybersecurity
Here is a case study with age and gender prediction in a chat room. One person's username, lonelygirl, suggests that she is a girl. This person also claims to be as old as the conversation partner, a 15-year-old teen. However, text analysis predicts that the person writes like an adult male (25+). The moderators should examine this conversation more closely.
Customization. We offer a number of tailored services, for example for cybersecurity (online harassment, hate). If you need a custom solution for a specific domain or language, feel free to contact us! If it doesn't exist, we'll help you build it.
Live demo
Try it out! Enter your text and tap ‘Analyze’. More text is better – a single word like ‘wow’ is harder to classify. The confidence score represents the average predictive accuracy: 80% = 8/10 correct predictions.
If you see a result that's wrong, you can help the AI get smarter. Simply tap the result field to correct it.