The First ESL Grammar Checker

Origin

The First and Best ESL Grammar Checker

The Virtual Writing Tutor is a free online English as a Second Language (ESL) grammar checker. Virtual Writing Tutor references over 5100 error detection rules developed by analyzing the errors of low-intermediate Francophone college (CEGEP) students' writing in English. Thanks to the generous support of Bokomaru Publications and with the help of a programmer,  VirtualWritingTutor.com was launched on April 5th, 2012.

Goal

Our goal from the outset with the Virtual Writing Tutor has been, above all, to enhance ESL pedagogy. Since one of the best ways to learn the structure of English is by writing in English, we figure that good pedagogy will be served through the development of a tool that supports teachers in their efforts to get students to write more. Automating corrective feedback could reduce the amount of time teachers spend giving corrective feedback to their students, and therefore should allow teachers to set more writing assignments for their students.

Here is the logic. The amount of writing college students do in a course is usually constrained by the amount of time a teacher has for evaluation and feedback, and, as any teacher will attest, most students are loath to write a multi-paragraphed, structured text unless it "counts." However, when an assignment counts, there is also the expectation of corrective feedback on grammar, word choice, and punctuation on one or more drafts.

I cannot attest for other countries or provinces, but in the Quebec junior college system (a.k.a. CÉGEP system), teachers routinely teach between 125-150 students per semester. Since each writing assignment will normally result in a minimum of 5-10 minutes of underlining, error-coding, or explicit correction per student, each writing assignment adds 10-25 hours of correction time to a teacher's otherwise busy week.

125 Ss x 5 min = 625 min = 10 h 25 min

150 Ss x 10 min = 1500 min = 25 h

With such an impact on workload, teachers usually do not assign more than one or two 350-500 word writing assignments per semester. Automating corrective feedback on errors, should therefore free teachers from the most time-consuming element of writing evaluation and allow them to set additional writing assignments each semester. More assignments will lead to more writing practice, more corrective feedback, and better learning.

Good ESL pedagogy is also served when a student receives corrective feedback in a timely fashion. The very fastest a teacher can provide hand-coded corrective feedback to a class is the week after they have handed in an assignment. Automatic corrective feedback provided online through a website such as the Virtual Writing Tutor can provide at least as much or substantially more corrective feedback in about one second. That's 144000 times faster!

In short, our goal for the Virtual Writing Tutor is to be non-trivial.

Theory

Underpinning the development of a grammar checker for English-Second-Language learners is the assumption that second language writing errors are both predictable and recurrent. Any teacher who has spent more than a year teaching writing can tell you that they see the same errors again and again from the same student and across groups of students, year after year.

Errors are understandable when we look at their causes. The most frequent errors in my students' writing tend to originate as translations of structures and sequences in the learner's first language. An example of one such error is the word choice error I have 18 years old for I am 18 years old, and the word order error She is a woman very popular for She is a very popular woman. Such transfer errors are very common and may need to be corrected many times before the learner is able to eliminate them from his or her writing.

Another high frequency type of error comes from learners simply not knowing the correct form, so an incorrect form is used in its place. That seems to be what is going on in sentences such as We did a lot of sandwiches.French employs faire in contexts where English uses either do or make. The problem is that when earners do not have an extensive knowledge of word collocations, they will guess between the two English forms. The learner will guess correctly sometimes and incorrectly other times.

A third type of error comes from the learner using a false analogy. The learner may know how to write a sentence such as It's big or It's far, but then he or she will falsely assume the sentence It's depend is equally correct because the sentence shares the first word it. Obviously, the learner is not yet aware that big and far are adjectives while depend is a verb and needs to be conjugated.The learner has yet to analyze enough English to know the difference.

A forth type is the false belief. Sometimes learners think they have understood something about English when in fact they have got the wrong idea. For example, one student wrote, You can describe myself as impulsive. What he meant was You could describe me as impulsive. The writer seems to have concluded that myself is an object pronoun.

Other errors are recurrent but are not specific to second language learners. They are due rather to overly rapid typing, inattention or fatigue. An extra keystroke can result in a typo, and a distraction can cause you to type a word twice. While an online spell/grammar checker can be useful in helping locate these recurrent errors, they can occur for all writers writing in either their first language or second language. Here we might label these mistakes rather than errors. Simply having a classmate proofread a text will be enough to eliminate them. No specialized knowledge is required.

While all these errors are recurrent, they are not entirely predictable, are they? We cannot always be sure what forms learners will bring over from their first language because of the principle of homoiophobia. Nor can we ever know with absolute certainty what learners have forgotten from prior instruction or what false analogies and what false beliefs they will come out with.However, the past predicts the future in a probabilistic way. In other words, the more frequently an error has occurred in the past, the more likely we can expect the same error to be committed in the future. 

Method

Since ESL writing errors tend to be somewhat recurrent in L2 writing, we assume that the best way to develop error detection rules is from authentic learner texts. One obvious place to start looking for high frequency errors is in learner corpora and our own learners' writing assignments. Of course, we would like our system to be useful to learners all over the world, so we equipped the Virtual Writing Tutor with a script to capture text submitted to the system and store it in a database for later rule development and refinement.

Some teacher-intuition comes into play as we review these texts for errors. High priority errors for us are those that relate to our lessons on English verb tense, aspect, and prepositions, and the error correction practice tasks we given them.

However, we do not rely on teacher-intuition alone to validate our rules. Collocations are checked in a 3+million word native speaker corpus in an attempt to create reliable rules and avoid false positives. Two things have become apparent in the process. First of all, 3 million words is not enough to capture the range of expressions and communicative functions that English has to offer. The corpus often comes up short when queried with what seems to me to be a common enough expression. Secondly, memory or teacher-intuition is not as reliable as you might think. Try it. You might be convinced--as I was--that to and said can never appear in sequence because infinitive structures must contain to + the base form of a verb, but a concordancer can show you an authentic context where to + said can occur. Creating an error-detection rule that identifies to + said would result in a false alarm, something we try hard to avoid.

Additional Features

The Virtual Writing Tutor is more than just an error detector. It is also an instant curriculum. When a learner submits his or her text with errors in it, the system returns corrective feedback messages and links to related online error correction activities. The idea is that motivated learners will be able to follow those links to individual activities and develop error-correction as a skill. For an example of the range of errors the system can detect and the type of individualize instruction the system can provide, see this click here. To see a random selection of sentences with errors, visit this random error-correction activity and follow the directions. 

Limitations

Whenever we attempt to show off the system to other teachers, problems become apparent.Teachers have two expectations. One is that the system will detect all errors in a text. The second is that the system will not detect any errors in a "correct" piece of writing. By reviewing the database of captured texts, we are able to recognize these tests as the come along. Telltale signs are overly simplistic sentences that contain little else but a single agreement or tense error. For example, He work or He work yesterday are recent examples. Then there is always some attempt at feeding the system nonsense, like He work work. Here is a recent example of a test a teacher set for our system and how it bombed.

False positives are a real concern.  Recently, I was shocked to discover that the sentence, I have to leave at... produced a false alarm. Koreans and Chinese have the family name I or Lee or Rhee, so when the Virtual Writing Tutor detected a singular proper name before a verb without any inflection, it returned the corrective feedback message that I have to should be I has to. I quickly added the exception. Needless to say, the Virtual Writing Tutor is a work in progress.

Finally, a more serious concern is that we will not be able to provide much help to advanced learners. That is a bridge we will cross once beginner and intermediate errors have been more fully dealt with. Stay tuned.

Conclusion

The Virtual Writing Tutor is a work in progress, but the goal is sound. Month after month, we continue to develop rules based on our students's errors. With the help of learners and teachers who alert us to missed errors and false alarms and at the rate we are going, in a year or two, the system will begin to compare favourably with the range of errors a human teacher can correct--only quicker. When that happens, we will have achieved our goal of creating a non-trivial pedagogical power-tool.

The 10 Best Grammar Checkers

I have added VirtualWritingTutor.com to an existing List.ly list. The link is here:

Best Grammar Checkers

To help you to consider the merits of the various systems in terms of addressing  English as a Second Language errors, I would like to offer examples of sentences selected from my students' writing. The best grammar checker on the list is Grammarbase with ten votes, so I submitted the sentences below to Grammarbase.com and received the message here to the right: "Congratulations! There are no critical mistakes in your writings." Judge for yourself.

After that, I had one of those bad day, everything badly begins this day. 

My school schedule is simple, everyday I start at 9 hr.

My last class of English it was in 2010.

I am studying here since two years.

I am in this program for 2 years

His mother took us there when we are young. 

I submitted this longer text, a text recently contributed by a user, and Grammarbase indicates that my text is genius level writing.

One day, I traveled to Ushuaia, but I had no idea than I would to find. I come to plane, I traveled four hours, and I arrived in Buenos Aires after, I come the second plane (the small and noisy plane) and I arrived in End of the World as the called the Ushuaia. I stayed wonder! The airport was built of wood and glass. The landscape was beautiful! The mountains and the sea together were amazing. I took the taxi and I arrived in the downtown. The city is as postcard. The homes had the European style and the diversity colours. The persons were friendly. There many activities for to make. I enjoyed a lot. The activity which I very loved was the track in the Park National. The thing very interesting was looked the evening at twenty three hours!

Please, try the samples given above on http://VirtualWritingTutor.com and then visit the List.ly list and decide which is really the best grammar checker for ESL students.

Virtual Writing Tutor Video Tour (en français)

Présentation du Virtual Writing Tutor : 

un correcteur web adapté à l'anglais langue seconde

Lien: Première édition sur Le Monde en Images

Le professeur d’anglais langue seconde, Nicholas Walker, du Collège Ahuntsic, présente l’outil web que il a créé dans le but d’assister ses étudiants francophones dans leurs travaux de rédaction. Il montre comment, du site web Virtual Writing Tutor, l’étudiant peut soumettre son texte à la correction. Il peut alors faire ses choix, à partir de suggestions de correction, et accéder à un relevé des fautes, incluant les statistiques de fréquence et les éléments de rétroaction corrective (emplacement et explication de la faute, suggestions de correction, lien vers une ressource en ligne, s’il y a lieu). L’étudiant peut imprimer son texte corrigé et y joindre le relevé.

 On peut en connaître davantage sur la démarche pédagogique de l’enseignant en faisant la lecture du récit de Profweb : « Le Virtual Writing Tutor : un correcteur web adapté à l’anglais langue seconde ». Nicholas Walker y fait valoir le fait que l’on peut varier les stratégies d’apprentissage de l’anglais langue seconde avec cet outil en intégrant un processus de révision dans les habitudes de rédaction des étudiants. L’outil permet d’augmenter la fréquence des évaluations et la précision des rétroactions tout en offrant une assistance à la rédaction tenant compte des principales faiblesses des étudiants francophones.

(Merci à Émilie Lavery !)

A Grammar Checker Fails the Test

The Virtual Writing Tutor Bombs Badly

A fellow ESL teacher recently sent me this message on Facebook about the VirtualWritingTutor, an English grammar checker for ESL learners.

I tried the VirtualWritingTutor with a text I ask my students to edit (English 100). It missed a lot of mistakes - especially with the present continuous verb forms. I think that you'll need to add even more error detection rules! 

Here's the text I tried out:

You have a pet? I do. I have a cats. It have five year old. I buy the animal for my childs. My son and daughter are very happy when they got Tommy. Tommy is the cat’s name. My son love the cat very much. He brings the cat in her bed at night and the cat sleep there. He is feeding the cat every day. Sometimes my daughter feed the cat. Last week, the cat go outside and don’t come back for two day. We not closing the door. Usually, we always close the door. This weekend, my friend is feed the cat because we going on a trip. We go at New York.

The Virtual Writing Tutor--a work in progress

Here was my reply

Thank you very much for this. Obviously, there are many more rules to write. The rules I am writing are based on authentic errors written by my own Francophone 100B and 101A students and based on the two student corpora on Lextutor.com . For some reason, the errors teachers submit to test the system almost always catch it off guard. I am not sure if it is how teachers reformulate errors from memory, the fact that different writing tasks produce a different range of errors, or just the sheer chaos of L2 writing.

In response to the text you posted, I have added a couple dozen rules to the system. Instead of catching 4 errors and throwing up false alarms, the system will now catch 17 errors in your text. Some errors I won’t try to catch simply because they involve understanding sentences that come before and after. This system is set up to look for overt sentence-level errors.

New Error Detection Rules Added

You have a pet?

I added a rule to catch the lack of an auxiliary with “You have a pet?” and a second rule for “You have pets?”

It have five year old.

I had a rule to catch “It have five years old” and many variations but not “it have five year old.” I modified the rule to catch the extra variation.

I buy the animal for my childs.

I created an exception for Childs in the proper name capitalization rule, reasoning that high frequency pluralization errors are more of a problem than low frequency family name capitalization errors. I will have to create a separate rule for the family name Childs. I created a rule to catch “childs” with the exception that it should be ignored if “childs” appears before another noun. In such a case, it will assume that the learner needs a possessive form. I won’t try to catch the “I buy” tense error. it is too broad of an assumption to expect all storytelling to use the past tense when dramatic stories can use the present simple also. More to the point, I have been focusing on sentence-level errors only.

My son and daughter are very happy when they got Tommy.

I created a rule to compare the verb at the beginning of the sentence with the verb at the end of the sentence. I have 36 of these rules so far. I’ll need many more, I figure.

My son love the cat very much.

I created an additional condition for the 3rd-person rule. It is easier to teach than to program because of the danger of false alarms. For one, you have to add an exception for the subjunctive. Also, we need an exception for all past forms, past participles, modals and infinitives that look like present simple forms.

I fixed the adjective order false alarm for “cat very much” by adding an exception that looks for a determiner before “cat”.

He brings the cat in her bed at night and the cat sleep there.

Now the system catches the third person singular error.

He is feeding the cat every day.

The system was able to catch “Every day, he is feeding the cat”. It is easier to program the system to catch aspect errors when the adverb of frequency is at the beginning or in the middle of the sentence. I have nevertheless added the new rule.

Sometimes my daughter feed the cat.

The reason the system didn’t catch “feed” is that it is also the low-frequency but nevertheless past tense of verb “to fee” which is excluded as being potentially correct. I had to create a separate rule for “feed”.

Last week, the cat go outside and don’t come back for two day.

This one was tough because of the two verb errors, but I have cracked it. I’ll need to develop more rules along this line.

This weekend, my friend is feed the cat because we going on a trip.

This was helpful for me. I set up a new group of rules to deal with future plans and arrangements. The second part was simply a case of catching a missing auxiliary after “because,” a case that is very unlikely to be a question with auxiliary-fronting

Stumping the Virtual Writing Tutor

Do you have any error-correction activities or sentences that stump the VirtualWritingTutor? If you do, please share them by adding them in a comment below. I will certainly add any missing rules in short order.

How to choose an English Grammar Checker

English Grammar Checkers Online

By Joel from Davis, CA, United States 
[CC-BY-2.0 (http://creativecommons.org/licenses/by/2.0)]
via Wikimedia Commons

There are a number of free online grammar checkers to choose from these days. Go ahead. Do a Google search using the search terms "English grammar checker" and you will see.

Some grammar checkers are free, some are pay-for-use, and some are just plain useless. If you teach English writing, you may be wondering which is the  right grammar checker for your learners.

To help you make the right choice, I put two English grammar checkers to the test to illustrate how each was designed with a different learner population in mind.

Wordiness and L1 Writing

Most grammar checkers have been designed to help first language (L1) writers detect and correct inadvertent mistakes and style errors. College-age writers struggling for the first time with the rigours of an academic style tend to fill their essays with redundant phrases, perhaps in an effort to achieve the minimum assigned word count. The sentence below is a great example of the kind of wordiness that abounds in L1 student writing. It seems to want to say something significant, but it contains far too many useless words. They obscure the idea that the sentence hopes to convey.

Due to the fact that access to internet resources are actually at the present time very easy to access in many places, the vast majority of users seek to have those kinds of devices that are most easy to carry around with them wherever they go.

I found this particular example in Dr. Kim Blank's wordiness list. The good doctor offers the following sentence as a worthy (less wordy) alternative:

 Because the Internet is available most places, users often prefer portable devices.

A good First Language (L1) grammar checker should help writers eliminate wordiness errors from their writing, In this regard, Grammark is a good L1 grammar checker. Plug the the first example into Grammark, and it  catches four wordiness errors. With its database of 973 wordy phrases, it is bound to help writers to excise much of the clutter from their writing. The Grammark grammar checker will not rewrite the sentence for you minus the errors. That you have to do yourself. As far as I know, no machine can rewrite your paper for you.  Rather,  grammar checkers like Grammark attempt to alert the writer to common errors inexperienced writers make in order to promote self-correction.

Grammar Checkers as Pedagogical Power Tools

Alerting a writer to errors achieves two important goals: 1) it helps the writer to eliminate errors from the draft he or she is writing at the moment, and 2) it helps the learner to eliminate those same errors from all future drafts by teaching the writer to recognize them in his or her writing. In other words, a grammar checker can help the learner both to revise bad writing and to learn the standards of good writing. As such, grammar checkers are power tools for teachers. Like other power tools, they don't replace master craftsmen, but they speed up the work.

Homophone Errors and L2 Writing

Catching homophone mix-ups is altogether a different problem for grammar checkers. You know them. They are everywhere in the comment sections of blogs and Youtube videos.  It's or its? Affect or effect? Your or you're?  Too, to, or two? There, they're, or their? We all confuse these homophones for each other on occasion, and so it is always best to have a second pair of eyes to proofread our writing whenever possible. It saves us from embarrassment. Consider these three (contrived) sentences.

Your crazy! Their are to many errors! Its effecting my mood.

Grammark, while undeniably helpful for first language writers of academic term papers and essays, is less of a champ at detecting homophone errors. It catches one (Their are → There are), but it misses the other four. Excellent in its own right, Grammark is not designed to catch grammatical errors of this kind. From what I can gather, it seems to be using a string matching technique, a method better suited for word choice or phrase choice errors. 

A Second Language Grammar Checker 

This is where a different kind of grammar checker really comes in handy. Less useful at detecting wordiness errors, the Virtual Writing Tutor sets its sights on overt grammar errors, the type of errors found in abundance in Second Language (L2) writing. Created to serve a different learner population, the Virtual Writing Tutor excels at catching homophone errors. Feed the same text into the Virtual Writing Tutor's text box, hit "Check grammar," the system returns five warnings, correctly identifying all the homophone errors. Here is what it says:

You wrote: Your crazy! Their are to many errors! Its effecting... Feedback: Did you mean 'you're crazy' or 'you are crazy'? 'Your' is a possessive adjective like 'his' or 'her'. You wrote: Your crazy! Their are to many errors! Its effecting my mo... Feedback: It looks as though you have confused 'their' and 'there'. Did you mean 'there'? You wrote: Your crazy! Their are to many errors! Its effecting my mood. Feedback: Remember that 'to' is a preposition and 'too' is an adverb meaning 'excessively'. Did you mean 'too many', as in 'excessively many'? You wrote: Your crazy! Their are to many errors! Its effecting my mood. Feedback: You probably meant 'It's' not 'its'. The word 'its' is a possessive adjective like 'his' and 'her', but 'it's' is a contraction of the pronoun 'it' and the verb 'is'. You wrote: ...ur crazy! Their are to many errors! Its effecting my mood. Feedback: You probably meant 'It's affecting my...'. The verb 'to effect' means to bring about a change. The verb 'to affect' means to have an effect on something.

Take-home Message

The take-home message here is simple. Grammar checkers specialize in catching certain errors because they are aimed at different learner populations. The Virtual Writing Tutor is great at catching homophone errors because it is primarily a Second Language grammar checker. It specializes in catching the kinds of errors that writers of English as a Second Language tend to make, a learner population that is typically under-served by systems designed to catch first language writing errors.

Thoughts, comments, reflections? Leave a comment.