- What is the goal of this website?
We try to provide the service as a writing tool to assist learners to improve their writing. Especially, we focus on "COLLOCATION", which is not only the common error in the learner's writing but also a hallmark of one's language proficiency.
- Is the service free?
Yes, it's definitely FREE! (Update 2010/06/05: We turn to Google Adsense now to support our website. No charge from users!)
- How do you come up with this idea?
We notice that one of the main reasons for learners to misuse collocations is mostly from the interference of their mother tongue. For example, in Chinese, "eating" is associated to "medicine" more often. Thus, for the collocation "______ medicine", learners who speak Chinese might directly translate the action to be "eating" but not "taking".
- How does it work?
We hypothesize that most wrong collocations might share the same translation with the correct counterparts. And we proposed that we can use the translations from dictionaries as clues to map back the correct collocation as suggestions.
- What kind of miscollocation problem will be handled?
We currently provide the service to handle those miscollocations resulted from the misused verbs in the V+N collocations. And we provide related suggestions manily concerning the verb, for example "release album" for "*publish album". We also provide example sentences for learners to see the usages in the authentic texts..
- Why does the system only provide suggestions concerning the verb?
Based on the literature, the verb is often the hardest part for learners to master in collocations. We thus would like to start with the research on verb., and hopefully extend our research to other types of miscollocations.
- Why can I not input more than 2 words?
We know this feature is long requested by our users. And we are currently working on a new service which can accept a longer query. We hope we can release this feature as soon as possible.
- What's your future plan?
We not only would like to handle other types or collocations, such as A+N or V+PREP+N ones, but also we plan to eliminate the limit on the length of users' query. We wish soon we can let users to input a full text as queries or provide the contextual editing function as what most existing spelling checkers do now.
