Anacode MarketMiner Simplifies China
via Automated Market Research
Anacode believes market research should be ready for the 21st century
That’s why we build MarketMiner, a text analytics software that cuts costs, time and complexity in market research – and gets you ready for taking better decisions based on the the authentic, intrinsic voice of your customers.
Anacode is driven by a team of professionals in the fields of NLP (Natural Language Processing), software development and China business. We develop advanced data and text analytics solutions which help Western, non-Chinese customers in their quest towards an independent and unbiased understanding of the Chinese market. Our technology comprises algorithms for data collection and analysis which extract fine-grained, statistically relevant insights from various types of customer and market feedback. By putting our technology at your service, we save you the cost, time and often frustration associated with field and desk research on the Chinese market.
In order to help you grow your China business, we offer technology-based solutions in the following areas:
MarketMiner is Anacode’s proprietary application for end customers who wish to get a solid understanding of the Chinese market. The backend performs a classical pipeline with three steps: online data collection, data analysis and aggregation. The frontend, designed in English, allows to formulate your research questions by products, product features and customer segments. The analysis and aggregation are fully transparent – you have the possibility to trace back your result via intermediate aggregation and analysis steps back to the raw data.
With the rise of Big Data, text analytics is showing a constant upheaval in the past years.
What is different about Anacode’s technology? Human language instead of words
All our development builds on a strong linguistic basis which includes carefully classified lexica and ontologies, linguistically annotated training data, full syntactic parsing as well as a logical interpretation language based on compositional semantics.
We KNOW what we analyze, and use this knowledge to create a linguistically adequate training space. The data mining algorithms achieve a higher accuracy than traditional approaches centered around simple collections of keywords.
Each version of the software treats a specific domain. We use language-specific grammars and specialized industry lexica to achieve an accurate analysis. At the same time, we build on principles of multilingual engineering so as to fully support the globalisation efforts of our clients. We know where to look for differences and similarities to streamline development while making appropriate customisations and thus maintaining a high per-case quality standard.
Depth of analysis
Our algorithms combine linguistic and statistical methods and thus allow to pull more detailed information from our data.
For example, traditional sentiment analysis positions an opinion on a positive/negative scale; thus, a sentence like “The seats are old-fashioned” will be analysed as a negative opinion.
MarketMiner decomposes and “explains” the statement by telling that it is the visual appearance of the seats which leads to a negative evaluation, and more specifically a negative evaluation in terms of fashionability.
Anacode’s Web&Text API offers data collection and analytics tailored to the Chinese web and language, allowing China-facing businesses to extract maximal value from customer and market feedback.
The API is perfectly suited for users who don’t speak Chinese and want a quick, hands-on start with “native” data. Get started with your own analytics pipeline in four steps:
- Register for a free account and upgrade when needed.
- Get data:download our web data for your domain, or extract your own web data using our scraping call.
- Analyze: extract categories, concepts and sentiments using our Natural Language Processing.
- Visualize and integrate the results which are returned in English – using your own applications or our Anacode Toolkit.
Not sure how Anacode can help you leverage the power of Chinese data? Request a live demo – we will be happy to walk you through our API and guide you in its proper usage for your specific requirements.
Market research has long relied on reactive means of data gathering, such as questionnaires or focus groups.
With the wide-spread use of social media, millions of comments about customer opinions and feedback regarding products and brands are available.
However, before using this ‘wisdom of the crowd’ as a source for marketing research, several challenges have to be tackled: the sheer volume of posts, their unstructured format, and the dozens of different languages used on the internet. All of them make automated usage of this data challenging.
In this paper, we draw on dashboard design principles and follow a design science research approach to develop a framework for search, integration, and analysis of cross-language user-generated content.
With ‘MarketMiner’, we implement the framework in the automotive industry by analyzing Chinese auto forums.
The results are promising in that MarketMiner can dramatically improve utilization of foreign-language social media content for market intelligence purposes.