Building Your Own Natural Language Processor - Tokenize

Now that we have sentences we need to break it into words. This phase is called "Tokenize".

Tokenize: This is the process of taking each sentence and separating it into "words" or tokens.

For a basic provider you can do a split into words. These functions can be found in the CSHARP.Text repository but I have placed them here as well. 

        /// <summary>
        /// Splits a string into its words for manipulation
        /// </summary>
        /// <param name="toSplit">String to split into words</param>
        /// <returns></returns>
        /// <remarks>Uses default values to split words</remarks>
        public List<string> SplitStringIntoWords(string toSplit)
        {
            return SplitStringIntoWords(toSplit, new char[] { ' ', ',', ';', ':', '(', ')', '{', '}', '[', ']', '!', '.', '?' });
        }

        /// <summary>
        /// Splits a string into its words for manipulation
        /// </summary>
        /// <param name="toSplit">String to split into words</param>
        /// <param name="endOfWordToken"></param>
        /// <returns></returns>
        /// <remarks>v2.0.0.11 Strips string before splitting into words</remarks>
        public List<string> SplitStringIntoWords(string toSplit, char[] endOfWordToken)
        {
            var words = new List<string>();
            var splitBuffer = toSplit.Trim();

            while (string.IsNullOrEmpty(splitBuffer) == false)
            {
                string foundWord = GetBeforeOneOf(splitBuffer, endOfWordToken, "EXCLUDING");

                // only add word if not empty string.
                if (string.IsNullOrEmpty(foundWord) == false) words.Add(foundWord);

                splitBuffer = (foundWord == splitBuffer)
                    ? string.Empty
                    : GetAfterOneOf(splitBuffer, endOfWordToken, "EXCLUDING").Trim();
            }

            return words;
        }

In your provider you would call SplitStringIntoWords(sentence) for each of your sentences and then store the tokens on the Tokens property for the sentence.

            foreach(var sentence in Sentences)
            {
                sentence.Tokens.AddRange(SplitStringIntoWords(sentence.Sentence));
            }

For your purposes you may need to add more items types of characters to split or even swap out the functionality to split into words for another. For the basic provider included in ByoNlpQuickStart we simply use this.

For DAIN and DIANA I started with this but found that I had to adjust the code to support situations where these characters could be the end of a word or simply be part of a word. The first exception I hit was urls mentioned in text. This would split http://www.test.com so that http was its own word. The sentence one would split into "http://www" as one sentence and "test" as another and then "com" as another. This is not correct, of course.

I added a rules property that stored JSON rules. Then during Tokenize it would look for specific exceptions before parsing. If none existed it could simply do split using the above but if there were exceptions then it would run the pre-processing step for the rules that apply and then split into words and then run the post-processing steps for those same rules. This has worked so far for DAIN and DIANA.

If you are interested in learning more about building your own then reach out and let's discuss. We have created a ByoNlpQuickStart that can help you get started. I will be providing a that QuickStart and others to SitecoreDain subscribers soon. If you cannot wait email me at chris.williams@readwatchcreate.com and as a subscriber, I can release an early version to get you started now.

Comments

Post a Comment

Popular posts from this blog

Natural Language Processing

Another key concept to DAIN and DIANA is Natural Language Processing. There are a variety of libraries and algorithms that can convert a sentence or series of statements into their various figures of speech.  DAIN and DIANA use the provider model the same way we did for Language Detection. This allows you to swap out your NLP as well as use multiple with the same idea of providing weighting to determine which provider would be considered more important. Weighting can be a simple number weighting or could include simple or more complex rules like this provider works better for English than it does for french or if this sentence contains these keywords then use this provider over that one. Natural Language Processing Steps There are various phases in Natural Language Processing. This article explains the basic ones.  Some libraries will expose the various steps while others will group them together or have a single method to do all of them: Splitting: This is the proces...
Read more

Providers and Pipeline, Oh Why!!!

I have mentioned to a few people that DAIN and DIANA are built with providers and pipelines. The pipelines. Some providers are weighted while others run conditionally based on rules.  The question I get is often is why not just choose an algorithm and stick with it for simplicity.  There are a few reasons for that: There are 3 reasons to have different providers the first is Bias and the second is Context. Bias In the case of Natural Language Processing, you have heard a lot in the news about bias in AI and ML. Some of that is due to data which is the subject of another article but other times it is that the algorithm has been trained with data that is not specific to your domain and thus can produce bias that is not specific to your data. That is why it is so important to to use multiple algorithms to keep the other in check. You can weight one algorithm higher so it is most likely to produce the result but there are cases where it differs greatly from other algorithms t...
Read more

The Universal Content Processing Engine

I am starting to learn more about Digital Asset Management and Content Processing and automatically processing content to gather Metadata both structured and unstructured.  Through this I have come up with a new self-learning engine called the Universal Content Processing Engine. Whereas currently I take advantage of statically programmed logic on how to process a tweet, make calls to given API based on static pipelines, I am limited to how I am coded to a certain extent although some learning has been happening.  Through the use of UCP Engine, I will start with a core set of knowledge and will build upon that.  I will then ask my maker for more information on how to process different types of data. Sometimes it would be a configuration of an existing component and other times it will be another provider dynamically added to my system. To start I will process content from 3 sources: FILE, STRING, URL. Through supervised learning I will learn about more sources and ...
Read more