The sequence similarity algorithm originated to meet the following specifications:
- A true representation of lexical similarity – strings with smaller distinctions ought to be seen as becoming comparable. In particular, an important sub-string overlap should indicate increased level of similarity between the chain.
- A robustness to changes of term purchase- two chain that have the same statement, however in an alternate order, must thought to be being similar. However, if a person sequence is a random anagram in the characters within the more, it should (usually) be named dissimilar.
- Code freedom – the formula should operate not just in English, but also in many different dialects.
Answer
The similarity was computed in three procedures:
- Partition each string into a summary of tokens.
- Processing the similarity between tokens through the use of a sequence edit-distance formula (extension function: semantic similarity description by using the WordNet collection).
- Computing the similarity between two token listings.
You will find another discussion for your guide.
A much better similarity ranking algorithm for varying duration chain
Many thanks all for the help and guidelines.
Martin Xie [MSFT] MSDN society help | Feedback to you Get or demand signal test from Microsoft Please be sure you mark the responds as solutions when they let and unmark them when they give no assistance.
- Marked as answer by Martin_Xie Monday, September 26, 2011 8:48 in the morning
All responds
What exactly is your question,explain they much more certain,i have confused with our
As an instance “a_logfile.txt” and “logfile_a.txt” must certanly be very similiar and aswell “loga_file.txt” and “logfile.text” not “myText.txt” and “logfile.txt”
Whether or not it resolved your trouble,Please simply click “Mark like address” thereon blog post and “tag as Helpful”. Pleased Development!
Ok I check it out again 🙂
Better I do want to examine filenames and i need to get a portion amounts in just how similiar they might be. We do not determine if this really is feasible anyway.
For instance a filename “a_filename.txt” and “filename_a.txt” is very similiar for people but exactly how could I have the same consequences programmatically.
Another example filename “file_abc_.txt” and fil_abc_e.txt” normally similiar but yet again how to have the lead programmaticaly
This is certainly perhaps harder than it seems at first.
Look at http://en.wikipedia.org/wiki/String_metrics and adhere certain links.
Relation David R Every regimen at some point becomes rococo, following rubble. – Alan Perlis the only real legitimate dimension of signal high quality: WTFs/minute.
Welcome to MSDN Discussion Board.
This short article shows the answer about: tips Compute the similarity between two words/strings. The formula originated in C# and you can install the trial in.
The sequence similarity algorithm was developed to satisfy here specifications:
- A true representation of lexical similarity – strings with small variations must be recognized as being similar. In particular, a substantial sub-string convergence should indicate a higher level of similarity involving the strings.
- A robustness to modifications of word order- two strings which contain alike words, however in a separate purchase, must thought to be are comparable. However, if an individual string is just a random anagram regarding the figures within the more, then it should (usually) getting named dissimilar.
- Code liberty – the formula should operate not only in English, and in many different languages.
Solution
The similarity are computed in three steps:
- Partition each string into a list of tokens.
- Computing the similarity between tokens with a string edit-distance formula (expansion function: semantic similarity dimension with the WordNet collection).
- Processing the similarity between two token databases.
There is another conversation for the research.
An improved similarity position formula for adjustable length strings
Thank you all for your assistance and tips.
Martin Xie [MSFT] MSDN people help | Feedback to you see or demand laws Sample from Microsoft Please don’t forget to mark the replies as answers when they help and unmark all of them when they incorporate no assistance.
- Marked as address by Martin_Xie Monday, September 26, 2011 8:48 AM
We have authored a rule for my job to discover similar brands roughly from database.
first I made use of the DIFFERENCE(string1, string2)>=4 function of SQL host however it don’t help me because like whenever first name is “21” and second term was “21 leap road” the end result included two brands whereas demonstrably they did not even similar. therefore the outcome group of these types of a query contained over 700 prices that was very poor in this situation.
I quickly found a similar HUGE DIFFERENCE features for c# that was nearly the same as SQL version of that features. for instance they matched up the similarity of “asdcdfsdfgdsgdg” and “asdewwetqwetrwe” as Great that’s obviously not the case.
I then developed a category for this problem to obtain additional effective similarity between strings.
title of this class is StringCompare and let me reveal an overview of this lessons:
SOMETHING STRING EXAMINE?
StringCompare was a comparing tool for chain. Maybe not an ordinal comparison, but a member of family review that identifies just how much two strings is close or exactly how much not comparable.
By setting the favorable tradeoff beliefs you can acquire an effective evaluation for strings.
HOW TO USE:
Initial you will want to setup an instance of StringCompare with tradeoff values or default tradeoff prices.
You will find 4 standards which can be ready:
1. MinSimilarityLong:
This is basically the minimum appropriate amount of similarity between two chain that researching with StringCompare. This advantages is utilized for chain together with the duration of at the very least 8.
2. MinSimilaritybrief:
This is basically the minimal appropriate portion of similarity between two chain that comparing with StringCompare. This worth is employed for strings using the length below 8.
3. MaxToleranceLong:
This is the optimal appropriate percentage of endurance between two chain that contrasting with StringCompare. This appreciate is employed for chain aided by the amount of at the least 8.
4. MaxToleranceShort:
This is basically the optimum acceptable amount of endurance between two chain that evaluating with StringCompare. This benefits can be used for chain with all the duration below 8.
* once you’ve developed a case you can easily phone InstanceName.IsEqual (string1, string2) to ascertain the equivalence of two strings.
* think about that equivalence was in accordance with the minSimilarty and maxTolerance your arranged before.
* Consider that larger minSimilarity beliefs can lead to more restricted effects and vice versa.
* start thinking about that reduced maxTolerance beliefs will result in most limited effects and vice versa.
Eg:
Recent Comments