I am new to SageMaker. I have a large csv dataset which I would like labelled:
sentence_id
sentence
pre_agreed_label
148392
A sentence
0
383294
Another sentence
1
For each sentence, I would like a) a yes/no binary classification in response to a question, and b) on a scale of 1-3, how obvious the classification was. I need the sentence id to map to other parts of the dataset, and will use the pre-agreed labels to assess accuracy.
I have identified SageMaker GroundTruth labelling jobs as a possible way to do this. Is this the best way? In trying to set it up I have run into a few problems.
The first problem is I can't find a way to display only the sentence column to the labellers, hiding the sentence_id and pre_agreed_labels.
The second is that there is either single labelling or multi labelling, but I would like a way to have two sets of single-selection labels:
Select one for binary classification:
Yes
No
Select one for difficulty of classification:
Easy
Medium
Hard
It seems as though this can be done using custom HTML, but I don't know how to do this - the template it gives you doesn't even render
Finally, having not used mechanical turk before, are there ways of ensuring people take the work seriously and don't just select random answers? I can see there's an option to have x number of people answer the same question, but is there also a way to put in an obvious question to which we already have a 'pre_agreed_label' every nth question, and kick people off the task if they get it wrong? There also appears to be a maximum of $1.20 per task which seems odd.
Related
I have a question about the word representation algorithms:
Which one of the algorithms word2Vec, doc2Vec and Tf-IDF is more suitable for handling text classification tasks ?
The corpus used in my supervised learning classification is composed of a list of multiple sentences, with both short length sentences and long length ones. As discussed in this thread, doc2vec vs word2vec choice is a matter of document length. As for Tf-Idf vs. word embedding, it's more a matter of text representation.
My other question is, what if for the same corpus I had more than one label to link to the sentences in it ? If I create multiple entries/labels for the same sentence, it affects the decision of the final classification algorithm. How can I tell the model that every label counts equal for every sentence of the document ?
Thank you in advance,
You should try multiple methods of turning your sentences into 'feature vectors'. There are no hard-and-fast rules; what works best for your project will depend a lot on your specific data, problem-domains, & classification goals.
(Don't extrapolate guidelines from other answers – such as the one you've linked that's about document-similarity rather than classification – as best practices for your project.)
To get initially underway, you may want to focus on some simple 'binary classification' aspect of your data, first. For example, pick a single label. Train on all the texts, merely trying to predict if that one label applies or not.
When you have that working, so you have a understanding of each step – corpus prep, text processing, feature-vectorization, classification-training, classification-evaluation – then you can try extending/adapting those steps to either single-label classification (where each text should have exactly one unique label) or multi-label classification (where each text might have any number of combined labels).
I have a task to complete.
There are two types of csv files 4000+ both related to each other.
2 types are:
1. Country2.csv
2. Security_Name.csv
Contents of Country2.csv:
Company Name;Security Name;;;;Final NOS;Final FFR
Contents of Security_Name.csv:
Date;Close Price;Volume
There are multiple countries and for each country multiple security files
Now I need to READ them do some CALCULATION and then WRITE the output in another files
READ
Read both the file Country 2.csv and Security.csv and extract all the data from them.
For example :
Read France 2.csv, extract Security_Name, Final NOS, Final FFR
Then Read Security.csv(which matches the Security_Name) and extract Date, Close Price, Volume
Calculation
Calculations are basically finding Median of the values extracted which is quite simple.
For Example:
Monthly Median Traded Values
Daily Traded Value of a Security ... and so on
Write
Based on the month I need to sort the output in two different file with following formats:
If Month % 3 = 0
Save It as MONTH_NAME.csv in following format:
Security name; 12-month indicator; 3-month indicator; FOT
Else
Save It as MONTH_NAME.csv in following format:
Security Name; Monthly Median Traded Value Ratio; Number of days Volume > 0
My question is how do I design my application in such a way that it is maintainable and the flow of data throughout the execution is seamless?
So first thing. Based on the kind of data you are looking to generate, I would probably be looking at moving this data to a SQL db if at all possible. This is "one SQL query" kind of stuff. And far more maintainable than C++ that generates CSV files from CSV files.
Barring that, I would probably look at using datamash and/or perl. On a Windows platform, you could do this through Cygwin or WSL. Probably less maintainable, but so much easier it's not too much of an issue.
That said, if you're looking for something moderately maintainable, C++ could work. The first thing I would do is design my input classes. Data-centric, but it can work. It sounds like you could have a Country class, a Security class, and a SecurityClose class...or something along those lines. You can think about whether a Security class should contain a collection of SecurityClosees (data), or whether the data should just be "loose" and reference the Security it belongs to. Same with the Country->Security relationship.
Once you've decided how all that's going to look, you want something (likely a function) that can tokenize a CSV line. So "1,2,3" gets turned into a vector<string> with the contents "1" "2" "3". Then, each of your input classes should have a constructor or initializer that takes a vector<string> and populates itself. You might need to pass higher level data along too. Like the filename if you want the security data to know which security it belongs to..
That's basically most of the battle there. Once you've pulled your data into sensibly organized classes, the rest should come more easily. And if you run into bumps, hopefully you can ask specific design or implementation questions from there.
I am looking for a good approach using python libraries to tackle the following problem:
I have a dataset with a column that has product description. The values in this column can be very messy and would have a lot of other words that are not related to the product. I want to know which rows are about the same product, so I would need to tag each description sentence with its main topics. For example, if I have the following:
"500 units shoe green sport tennis import oversea plastic", I would like the tags to be something like: "shoe", "sport". So I am looking to build an approach for semantic tagging of sentences, not part of speech tagging. Assume I don't have labeled (tagged) data for training.
Any help would be appreciated.
Lack of labeled data means you cannot apply any semantic classification method using word vectors, which would be the optimal solution to your problem. An alternative however could be to construct the document frequencies of your token n-grams and assume importance based on some smoothed variant of idf (i.e. words that tend to appear often in descriptions probably carry some semantic weight). You can then inspect your sorted-by-idf list of words and handpick(/erase) words that you deem important(/unimportant). The results won't be perfect, but it's a clean and simple solution given your lack of training data.
I'm developing a Weka app like Akinator by using the j48 method.
Sample:
http://jbossews-vdoctor.rhcloud.com/doctor
The following is the app's table definition and sample data
qa means question id(Please refer the master which can be set by user) + answer(1:Yes, 2: I don't know, 3: No).
1 line per 1 question & answer.
id,qa,class
A,13,1
A,23,1
B,13,2
B,21,2
The point is to find a way to select the question which can maximize the entropy.
Currently this app is regarding first node id of decision tree as the best question.
And then it narrows down the options by this elimination way.
But the accuracy was too bad to run correctly so I'd like to improve it.
I noticed that the qa column was identified as numeric so it could not build the correct decision tree.
I am confused what I should do for improvement. Dataset? Table definition? Logic?
This is quite a broad question that you are asking, and without code or a clear understanding of the problem it is quite difficult to answer, but I'll give some tips for improvement:
Table Definition
What may have made more sense here is to have an attribute for each question, instead of using a single instance per question. For Example, instead of id, qa and class, you could have A, B, C, D, E, F and Disease. (I believe there were six questions, and naming each attribute would be recommended instead of A-F)
Dataset
You will need at least as many cases as there are diseases, if not more for defining multiple subsets of the problem space for the same disease. There are likely cases where some questions are irrelevant or missing, and the model may need to handle such situations.
Logic
In such a case, you might be able to do the questionnaire by starting with the root node and asking questions until you reach the estimated class. This way, you can ask from node to node until a class is reached.
I hope this helps in improving your existing model.
NOTE: I tried your questionnaire and answered No to all of your questions, and I strangely ended up with Trichomoniasis. Perhaps there could be a 'No Disease' category for your training data also.
My nominal qa data is building such a decision tree by binary split.
actually this structure won't make sense because there is tree at only one side. When qa equal 23 it would be always '3' answer. It's irrational.
http://www.fastpic.jp/viewer.php?file=2693704973.jpg
You should first reformat your features to get all possible questions A,B,C,D... as binary features and your final answer (ie. what to guess) as target class if you want your tree to get a sequence of questions reaching to your answer. Your data will certainly be sparse (many questions without data/answer).
By the way, a binary tree is not the right ML structure and algorithm to build an Akinator like or 20Q/Guess-who. Please look some suggestions here: https://stats.stackexchange.com/questions/6074/akinator-com-and-naive-bayes-classifier
I am relatively new to the data mining area and have been experimenting with Weka.
I have a dataset which consists of almost 8000 records related to customers and items they have purchased. 58% of this data set has missing values for the "Gender" attribute.
I want to find the missing gender values based on the other data I do have.
I first thought I could do this using a classifier algorithm in Weka using a training set to build a model. Based on examples I saw online, I tried this with pretty much all the available algorithms available in Weka using a training set that consisted of 60-80% of the data which did not have missing values. This gave me a lower accuracy rate than I wanted (80-86% depending on the algorithm used)
Did I go about this correctly? Is there a way to improve this accuracy? I experimented with using different attributes, different pre-processing of the data etc.
I also tried using the ReplaceMissingValues filter on the complete dataset to see how that would handle the missing values. However, it just changed all the missing values to "Female" which obviously cannot be the case. So I'm wondering also wondering if I need to use this filter in my situation or not.
It sounds like you went about it in the correct way. The ReplaceMissingValues filter replaces the missing values with the most frequent of the non-missing values I think, so it is not what you want in this case.
A better way to get an idea of the true accuracy of your gender-predictor would be to use cross-validation instead of the training/test split (Weka has a separate option for that). 80-86% may seem low, but keep in mind that random guessing will only get you about 50%, so it's still a lot better than that. To try to get better performance, pick a classifier that performs well and then play with its parameters until you get better performance. This is likely to be quite labour-intensive (although you could of course use automated methods for tuning, see e.g. Auto-WEKA), but the only way to improve the performance.
You can also combine the algorithm you choose with a separate feature selection step (Weka has a special meta-classifier for this). This may improve performance, but again you'll have to experiment to find the particular configuration that works for you.