Skip to content

milahu/reverse-template-engine

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

25 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

reverse template engine

reverse-engineer a JSX template of many similar HTML files

similar projects

commercial

most commercial scraping services offer this feature

status

  • it can compare two files
  • it can compare multiple files
  • it can find arrays of similar items
  • it can find trees of similar items
  • it can find optional blocks (conditional blocks)
  • the resulting parser is aware of html syntax

what

a template engine

  • takes one template (page.html.tpl) and an array of data (page.json)
  • returns an array of texts (page.html)
               ┌────────────┐
Template  ────►│            │
               │  Template  │
               │            │
               │   Engine   │
  Data[]  ────►│            ├────►  Text[]
               └────────────┘

a reverse template engine does the opposite

  • takes an array of texts (page.html)
  • returns one template (page.html.tpl) and an array of data (page.json)
               ┌────────────┐
               │            ├────►  Template
               │  Reverse   │
               │  Template  │
               │   Engine   │
  Text[]  ────►│            ├────►  Data[]
               └────────────┘

based on

common substrings

suffix tree

common pattern

learning algorithms

http://libalf.informatik.rwth-aachen.de/

Algorithm offline online target model
Angluin's L*   X DFA
L* (adding counter-examples to columns)   X DFA
Kearns / Vazirani   X DFA
Rivest / Schapire   X DFA
NL*   X NFA
Regular positive negative inference (RPNI) X   DFA
DeLeTe2 X   NFA
Biermann & Feldman's algorithm X   NFA
Biermann & Feldman's algorithm (using SAT-solving) X   DFA

Glossary

DFA: a deterministic finite-state automaton

NFA: a nondeterministic finite-state automaton

Offline learning: offline learning algorithms are learning algorithms that passively receive a set of classified data. Their goal then is to generalize this set of positive and negative words to some kind of explanation H (e.g. a DFA) which is in conformance with the input. aka non-supervised learning, passive learning.

Online learning: In contrast to offline learning algorithms, online learning algorithms are capable of actively asking certain kinds of queries to some teacher who is able to classify these queries. This ability lets them infer explanations for the underlying set of already classified words. aka supervised learning, active learning.

grammar inference

1.1 The Problem and Its Various Formulations

Let us start with the presentation of how many variants of a grammatical inference problem we may be faced with.

Informally, we are given a sequence of words and the task is to find a rule that lies behind it.

Different models and goals are given by response to the following questions:

  • Is the sequence finite or infinite?
  • Does the sequence contain only examples (positive words) or also counter-examples (negative words)?
  • Is the sequence of the form: all positive and negative words up to a certain length n?
  • What is meant by the rule: are we satisfied with
    • regular acceptor
    • contextfree grammar
    • context-sensitive grammar
    • other tool?
  • Among all the rules that match the input, should the obtained one be of a minimum size?

— Wojciech Wieczorek - Grammatical Inference. Algorithms Routines and Applications (2017)

Grammatical Inference

The problem of inducing, learning or inferring grammars has been studied for decades, but only in recent years has grammatical inference emerged as an independent field with connections to many scientific disciplines, including bio-informatics, computational linguistics and pattern recognition.

This book meets the need for a comprehensive and unified summary of the basic techniques and results, suitable for researchers working in these various areas.

In Part I, the objects of use for grammatical inference are studied in detail: strings and their topology, automata and grammars, whether probabilistic or not.

Part II carefully explores the main questions in the field: what does learning mean? How can we associate complexity theory with learning?

In Part III the author describes a number of techniques and algorithms that allow us to learn from text, from an informant, or through interaction with the environment. These concern automata, grammars, rewriting systems, pattern languages and transducers.

— Colin de la Higuera - Grammatical Inference. Learning Automata and Grammars (2010)

Information Extraction

2.4.2 Information extraction: automatic wrapper generation

The quantity of structured data available today due to the exponential growth of the World Wide Web introduces a number of challenges to scientists interested in grammatical inference. HTML and XML data appear as text, but the text is well bracketed through a number of tags that are either syntactic (HTML) and give indications as to how the file should be represented, or semantic (XML).

Here is a small piece from an XML file:

<book>
  <chapter>
    <name>Introduction</name>
    <length>25 pages</length>
    <description>
      Motivations about the book
    </description>
    <exercises>0</exercises>
  </chapter>
  <chapter>
    <name>The Data</name>
    <length>18 pages</length>
    <description>
      Describe some cases where the data is made of strings
    </description>
    <exercises>0</exercises>
  </chapter>
  <chapter>
    <name>Strings and Languages</name>
    <length>35 pages</length>
    <description>
      Definitions of strings and stringology
    </description>
    <exercises>23</exercises>
  </chapter>
</book>

Between the many problems when working with these files, one can aim to find the grammar corresponding to a set of XML files.

One very nice application in which grammatical inference has been helpful is that of building a wrapper automatically (or semi-automatically). A wrapper is supposed to take a web page and extract from it the information for which it has been designed.

For instance, if we need to build a mailing list, the wrapper would find in a web page the information that is needed. Obviously, the wrapper will work on the code of the web page: the HTML or XML file. Therefore, grammatical inference of tree automata is an obvious candidate.

Another feature of the task is that labelling the examples is cumbersome and can be noisy. The proposal is to do this on the fly, through an interaction between the system and the user. This will justify in part the rising interest in active learning methods.

— Colin de la Higuera - Grammatical Inference. Learning Automata and Grammars (2010)

Information Extraction from Structured Text

https://www.google.com/search?q=xml+information+extraction+automatic+wrapper+generation+grammatical+inference+of+tree+automata

Information extraction in structured documents using tree automata induction

Information extraction in structured documents using tree automata induction. R Kosala, JV Bussche, M Bruynooghe. 2002 - 60 quotes

A problem, however, in directly applying tree automata to tree-structured documents such as HTML or XML documents, is that the latter trees are “unranked”: the number of children of a node is not fixed by the label, but is varying.

There are two approaches to deal with this situation:

  1. The first approach is to use a generalized notion of tree automata towards unranked tree formalisms (e.g., [17,23]). In such formalisms, the transition rules are of the form δ(v, e) → q, where e is a regular expression over Q that describes a sequence of states.
  2. The second approach is to encode unranked trees into ranked trees, specifically, binary trees, and to use existing tree automata inference algorithms for inducing the tree automaton.

In this paper we follow the second approach, because it seems less complicated. An advantage is that we can use existing learning methods that work on ranked trees. A disadvantage is that we have to preprocess the trees before applying the algorithm.

17: C. Pair and A. Quere. D´efinition et etude des bilangages r´eguliers. Information and Control, 13(6):565–593, 1968.

23: M. Takahashi. Generalizations of regular sets and their application to a study of context-free languages. Information and Control, 27:1–36, 1975.

Relation extraction

https://medium.com/@andreasherman/different-ways-of-doing-relation-extraction-from-text-7362b4c3169e

Different ways of doing Relation Extraction from text

Relation Extraction (RE) is the task of extracting semantic relationships from text, which usually occur between two or more entities. These relations can be of different types. E.g “Paris is in France” states a “is in” relationship from Paris to France. This can be denoted using triples, (Paris, is in, France).

Information Extraction (IE) is the field of extracting structured information from natural language text. This field is used for various NLP tasks, such as creating Knowledge Graphs, Question-Answering System, Text Summarization, etc. Relation extraction is in itself a subfield of IE.

https://github.com/ekalgolas/Relation-extraction-using-Semantic-Web

We will process unstructured data from web (obtained by crawling some sample websites) by maybe: having a Apache SolR installation locally and manually feeding it web pages. We can use Stanford NLP API to extract semantics from the unstructured text. After we extract some semantics, we can construct a structured data format, probably RDF/XML/OWL and also have a visual representation of the graph data using Gruff.

https://www.slideshare.net/butest/information-extraction-from-html-general-machine-learning

search engines

  • apache lucene?
  • solr?
A Survey of Web Information Extraction Systems

A Survey of Web Information Extraction Systems. CH Chang, M Kayed, R Girgis, KF Shaalan, IEEE Transactions 2006 - 1200 quotes

4.2 Supervised WI systems

As shown in the left-bottom of Figure 5, supervised WI systems take a set of web pages labeled with examples of the data to be extracted and output a wrapper. The user provides an initial set of labeled examples and the system (with a GUI) may suggest additional pages for the user to label. For such systems, general users instead of programmers can be trained to use the labeling GUI, thus reducing the cost of wrapper generation. Such systems are SRV, RAPIER, WHISK, WIEN, STALKER, SoftMealy, NoDoSE, DEByE.

SRV is a top-down relational algorithm that generates single-slot extraction rules [8]. It regards IE as a kind of classification problem. The input documents are tokenized and all substrings of continuous tokens (i.e. text fragments) are labeled as either extraction target (positive examples) or not (negative examples). The rules generated by SRV are logic rules that rely on a set of token-oriented features (or predicates). These features have two basic varieties: simple and relational. A simple feature is a function that maps a token into some discrete value such as length, character type (e.g., numeric), orthography (e.g., capitalized) and part of speech (e.g., verb). A relational feature maps a token to another token, e.g. the contextual (previous or next) tokens of the input tokens. The learning algorithm proceeds as FOIL, starting with entire set of examples and adds predicates greedily to cover as many positive examples and as few negative examples as possible. For example, to extract the rating score for our running example, SRV might return rule like Figure 9(a), which says rating is a single numeric word and occurs within a HTML list tag.

https://www.cs.utexas.edu/~ml/rapier.html - ftp://ftp.cs.utexas.edu/pub/mooney/rapier - RAPIER is a bottom-up inductive learning system for learning information extract rules. It has been tested on several domains and performs comparably to or slightly better than other recent learning system for this task.

RAPIER also focuses on field-level extraction but uses bottom-up (compression-based) relational learning algorithm [7], i.e. it begins with the most specific rules and then replacing them with more general rules. RAPIER learns single slot extraction patterns that make use of syntactic and semantic information including part-of-speech tagger or a lexicon (WordNet). The extraction rules consist of three distinct patterns. The first one is the pre-filler pattern that matches text immediately preceding the filler, the second one is the pattern that match the actual slot filler, finally the last one is the post-filler pattern that match the text immediately following the filler. As an example, Figure 9(b) shows the extraction rule for the book title, which is immediately preceded by words “Book”, “Name”, and “”, and immediately followed by the word “”. The “Filler pattern” specifies that the title consists of at most two words that were labeled as “nn” or “nns” by the POS tagger (i.e., one or two singular or plural common nouns).

WIEN: Kushmerick identified a family of six wrapper classes, LR, HLRT, OCLR, HOCLRT, N-LR and N-HLRT for semi-structured Web data extraction [9]. WIEN focuses on extractor architectures. The first four wrappers are used for semi-structured documents, while the remaining two wrappers are used for hierarchically nested documents. The LR wrapper is a vector of 2K delimiters for a site containing K attributes. For example, the vector (‘Reviewer name ’, ‘’, ‘Rating ’, ‘’, ‘Text ’, ‘’) can be used to extract 3-slot book reviews for our running example. The HLRT class uses two additional delimiters to skip over potentially-confusing text in either the head or tail of the page. The OCLR class uses two additional delimiters to identify an entire tuple in the document, and then uses the LR strategy to extract each attribute in turn. The HOCLRT wrapper combines the two classes OCLR and HLRT. The two wrappers N-LR and N-HLRT are extension of LR and HLRT and designed specifically for nested data extraction. Note that, since WIEN assumes ordered attributes in a data record, missing attributes and permutation of attributes can not be handled.

WHISK uses a covering learning algorithm to generate multi-slot extraction rules for a wide variety of documents ranging from structured to free text [6]. When applying to free text, WHISK works best with input that has been annotated by a syntactic analyzer and a semantic tagger. WHISK rules are based on a form of regular expression patterns that identify the context of relevant phrases and the exact delimiters of those phrases. It takes a set of hand-tagged training instances to guide the creation of rules and to test the performance of the proposed rules. WHISK induces rules top-down, starting from the most general rule that covers all instances, and then extending the rule by adding terms one at a time. For example, to generate 3-slot book reviews, it start with empty rule “()()()*”, where each parenthesis indicates a phrase to be extracted. The phrase within the first set of parentheses is bound to the first variable $1, and the second to $2, and forth. Thus, the rule in Figure 10 can be used to extract our 3-slot book reviews for our running example. If part of the input remains after the rule has succeeded, the rule is re-applied to the rest of the input. Thus, the extraction logic is similar to the LR wrapper for WIEN.

https://dl.acm.org/doi/abs/10.1145/276305.276330 - NoDoSE — a tool for semi-automatically extracting structured and semistructured data from text documents. B. Adelberg, SIGMOD 1998

NoDoSE: Opposed to WIEN, where training examples are obtained from some oracles that can identify interesting types of fields within a document, NoDoSE provides an interactive tool for users to hierarchically decompose semistructured documents (including plain text or HTML pages) [23]. Thus, NoDoSE is able to handle nested objects. The system attempts to infer the format/grammar of the input documents by two heuristic-based mining components: one that mines text files and the other parses HTML code. Similar to WIEN, the mining algorithms try to find common prefix and suffix as delimiters for various attributes. Although it does not assume the order of attributes within a record to be fixed, it seeks to find a totally consistent ordering for various attributes in a record. The result of this task is a tree that describes the structure of the document. For example, to generate a wrapper for the running example, the user can interact with the NoDoSE GUI to decompose the document as a record with two fields: a book title (an attribute of type string) and a list of Reviewer, which is in turn a record of the three fields RName (string), Rate (integer), and Text (string). Next, NoDoSE then automatically parses them and generates the extraction rules.

SoftMealy: In order to handle missing attributes and attribute permutations in input, Hsu and Dung introduce the idea of finite-state transducer (FST) to allow more variation on extractor structures [10]. A FST consists of two different parts: the body transducer, which extract the part of the page that contains the tuples (similar to HLRT in WIEN), and the tuple transducer which iteratively extracts the tuples from the body. The tuple transducer accepts a tuple and returns its attributes. Each distinct attribute permutation in the page can be encoded as a successful path from start state to the end state of the tuple transducer; and the state transitions are determined by matching contextual rules that describe the context delimiting two adjacent attributes. Contextual rules consist of individual separators that represent invisible borderlines between adjacent tokens; and an inductive generalization algorithm is used to induce these rules from training examples. Figure 11 shows an example of FST that can be used to extract the attributes of the book reviews: the reviewer name (N), the rating (R), and the comment (T). In addition to the begin and end states, each attribute, A , is followed by a dummy state, A . Each arc is labeled with the contextual rule that enables the transition and the tokens to output. For example, when the state transition reaches to the R state, the transducer will extract the attribute R until it matches the contextual rules s<R, R > (which is composed of s<R, R >L and s<R, R >R ). The state R and the end state are connected if we assume no comment can occur.

STALKER is a WI system that performs hierarchical data extraction [11]. It introduces the concept of embedded catalog (EC) formalism to describe the structure of a wide range of semi-structured documents. The EC description of a page is a tree-like structure in which the leaves are the attributes to be extracted and the internal nodes are lists of tuples. For each node in the tree, the wrapper needs a rule to extract this node from its parent. Additionally, for each list node, the wrapper requires a list iteration rule that decomposes the list into individual tuples. Therefore, STALKER turns the difficult problem of extracting data from an arbitrary complex document into a series of easier extraction tasks from higher level to lower level. Moreover, the extractor uses multi-pass scans to handle missing attributes and multiple permutations. The extraction rules are generated by using of a sequential covering algorithm, which starts from linear landmark automata to cover as many positive examples as possible, and then tries to generate new automata for the remaining examples. A Stalker EC tree that describes the data structure of the running example is shown in Figure 12(a), where some of the extraction rules are shown in Figure 12(b). For example, the reviewer ratings can be extracted by first applying the List(Reviewer) extraction rule (which begins with “

    ” and ends with “
”) to the whole document, and then the Rating extraction rule to each individual reviewer, which is obtained by applying the iteration rule for List(Reviewer). In a way, STALKER is equivalent to multi-pass Softmealy [30]. However, the extraction patterns for each attribute can be sequential as opposed to the continuous patterns used by Softmealy.

DEByE (Data Extraction By Example): Like NoDoSE, DEByE provides an interactive GUI for wrapper generation [24], [25]. The difference is that in DEByE the user marks only atomic (attribute) values to assemble nested tables, while in NoDoSE the user decomposes the whole document in a top-down fashion. In addition, DEByE adopts a bottom-up extraction strategy which is different from other approaches. The main feature of this strategy is that it extracts atomic components first and then assembles them into (nested) objects. The extraction rules, called attributevalue pair patterns (AVPs), for atomic components are identified by context analysis: starting with context length 1, if the number of matches exceeds the estimated number of occurrences provided by the user, it adds additional terms to the pattern until the number of matches is less than the estimated one. For example, DEByE generates AVP patterns, “Name* Reviews”, “Name Rating”, “RatingText” and “

  • ” for book name, reviewer name, rating and comment respectively ( denotes the data to be extracted). The resulting AVPs are then used to compose an object extraction pattern (OEPs). OEPs are trees containing information on the structure of the document. The sub-trees of an OEP are themselves OEPs, modeling the structure of component objects. At the bottom of the hierarchy lie the AVPs that used to identify atomic components. The assemble of atomic values into lists or tuples is based on the assumption that various occurrences of objects do not overlap each other. For nonhomogeneous objects, the user can specify more than one example object, thus creating a distinct OEP for each example.

    4.3 Semi-Supervised IE systems

    The systems that we categorize as semi-supervised IE systems include IEPAD, OLERA and Thresher. As opposed to supervised approach, OLERA and Thresher accept a rough (instead of a complete and exact) example from users for extraction rule generation, therefore they are called semisupervised. IEPAD, although requires no labeled training pages, post-effort from the user is required to choose the target pattern and indicate the data to be extracted. All these systems are targeted for record-level extraction tasks. Since no extraction targets are specified for such systems, a GUI is required for users to specify the extraction targets after the learning phase. Thus, users’ supervision is involved.

    IEPAD is one of the first IE systems that generalize extraction patterns from unlabeled Web pages [31]. This method exploits the fact that if a Web page contains multiple (homogeneous) data records to be extracted, they are often rendered regularly using the same template for good visualization. Thus, repetitive patterns can be discovered if the page is well encoded. Therefore, learning wrappers can be solved by discovering repetitive patterns. IEPAD uses a data structure called PAT trees which is a binary suffix tree to discover repetitive patterns in a Web page. Since such a data structure only records the exact match for suffixes, IEPAD further applies center star algorithm to align multiple strings which start from each occurrence of a repeat and end before the start of next occurrence. Finally, a signature representation is used to denote the template to comprehend all data records. For our running example, only page pe2 can be used as input to IEPAD. By encoding each tag as an individual token and any text between two adjacent tags as a special token “T”, IEPAD discover the pattern “

  • TTTT TT
  • ” with two occurrences. The user then has to specify, for example, the 2nd, 4th and 6th “T” tokens, as the relevant data (denoting reviewer name, rating and comment, respectively).

    OLERA is a semi-supervised IE system that acquires a rough example from the user for extraction rule generation [32]. OLERA can learn extraction rules for pages containing single data records, a situation where IEPAD fails. OLERA consists of 3 main operations. (1) Enclosing an information block of interest: where the user marks an information block containing a record to be extracted for OLERA to discover other similar blocks (using approximate matching technique) and generalize them to an extraction pattern (using multiple string alignment technique). (2) Drilling-down/rollingup an information slot: drilling-down allows the user to navigate from a text fragment to more detailed components, whereas rolling-up combines several slots to form a meaningful information unit. (3) Designating relevant information slots for schema specification as in IEPAD.

    Thresher [33] is also a semi-supervised approach that is similar to OLERA. The GUI for Thresher is built in the Hay stack browser which allows users to specify examples of semantic contents by highlighting them and describing their meaning (labeling them). However, it uses tree edit distance (instead of string edit distance as in OLERA) between the DOM subtrees of these examples to create a wrapper. Then it allows the user to bind the semantic web language RDF (Resource Description Framework) classes and predicates to the nodes of these wrappers.

    DEPTA (Data Extraction based on Partial Tree Alignment): Like IEPAD and DeLa, DEPTA can be only applicable to Web pages that contain two or more data records in a data region. However, instead of discovering repeat substring based on suffix trees, which compares all suffixes of the HTML tag strings (as the encoded token string described in IEPAD), it compares only adjacent substrings with starting tags having the same parent in the HTML tag tree (similar to HTML DOM tree but only tags are considered). The insight is that data records of the same data region are reflected in the tag tree of a Web page under the same parent node. Thus, irrelevant substrings do not need to be compared together as that in suffix-based approaches. Furthermore, the substring comparison can be computed by string edit distance instead of exact string match when using suffix trees where only completely similar substrings are identified. The described algorithm, called MDR [38], works in three steps. First, it builds an HTML tag tree for the Web page as shown in Figure 14 where text strings are disregarded. Second, it compares substrings for all children under the same parent. For example, we need to make two string comparison, (b1, b2) and (b2, ol), under parent node

    , where the tag string node
      is represented by “
    1. ”. If the similarity is greater than a predefined threshold (as shown in the shaded nodes in Figure 14), the nodes are recorded as data regions. The third step is designed to handle situations when a data record is not rendered contiguously as assumed in previous works. Finally, the recognition of data items or attributes in a record is accomplished by partial tree alignment [39]. Tree alignment is better than string alignment for it considers tree structure, thus, reducing the number of possible alignments. The algorithm first chooses the record tree with the largest number of data items as center and then matches other record trees to the center tree. However, DEPTA only adds tag nodes to the center tree when the positions of the tag nodes can be uniquely determined in the center tree. For remained nodes, they are processed in the next iteration after all tag trees are processed. Note that DEPTA assumes that non-tag tokens are data items to be extracted, thus, it extracts not only the reviewer name, rating and comments, but also the labels “Reviewer Name”, “Rating”, and “Text” for page pe2 in our running example. Further, DEPTA is limited to handle nested data records. So, a new algorithm, NET, is developed to handle such data records by performing a postorder traversal of the visual-based tag tree of a Web page and matching subtrees in the process using a tree edit distance method and visual cues [40].

      Of the unsupervised WI approaches, one important issue is to differentiate the role of each token: either a data token or template token. Some assume that every HTML tag is generated by the template and other tokens are data items to simplify the issue (as in DeLa and DEPTA). However, the assumption does not hold for many collections of pages (therefore, IEPAD and OLERA simply leave the issue to distinguish between data and template tokens to the users). RoadRunner also assumes that every HTML tag is generated by the template, but other matched string tokens are also considered as part of the template. In comparison, EXALG has the most detailed tokenization method while more flexible assumption where each token can be a template token if there are enough tokens to form frequently occurring equivalence class.

      On the other hand, DEPTA conducts the mining process from single Web pages, while RoadRunner and EXALG do the analysis from multiple Web pages (While DeLa takes advantages of multiple input pages for data-rich section extraction and generalized pattern construction, it discovers C-repeat patterns from single Web pages.). The later, in our viewpoint, is the key point that is used to differentiate the role of each token. Thus, multiple pages of the same class is also used to discover data rich section (as in DeLa) or eliminate noisy information (as in [41]). Meanwhile, the adaptation of tree matching in DEPTA (as well as Thresher) also provides better result than string matching techniques used in IEPAD and RoadRunner. EXALG similarly does not make full use of the tree structure although the DOM tree path information is used for differentiating token roles. Finally, since information extraction is only a part of a wrapper program or information integration systems, additional tasks like page fetching, label assignment, and mapping with other web data sources are remained to be processed.

      Due to space limitation, we are not able to compare all researches here. For example, ViNTs [42] is a record-level wrapper generation system which exploits visual information to find separators between data regions from search result pages. However, the algorithm can be only applicable to pages that contain at least four data records. Another related approach that has been applied on Web sites for extracting information from tables is [43]. The technique relies on the use of additional links to a detail page containing additional information about that item. In parallel to the efforts to detect Web tables, other researchers have worked in detecting tables in plain text documents (such as government statistical reports) and segmenting them into records [44]. Since these approaches do not address the problem of distinguish data tokens from template tokens, we consider them as semi-supervised approaches.

    An XML-enabled data extraction toolkit for web sources

    An XML-enabled data extraction toolkit for web sources. Ling Liu, C. Pu, Wei Han, 2001

    http://maaz.ihmc.us/rid=1228288297429_1842851864_16163/An%20XML%20enabled%20data%20extraction%20toolkit.pdf

    Template Induction

    Discovering Textual Structures: Generative Grammar Induction using Template Trees

    Discovering Textual Structures: Generative Grammar Induction using Template Trees. Thomas Winters, L. D. Raedt, 2020

    https://arxiv.org/abs/2009.04530

    Latent Template Induction with Gumbel-CRFs

    Latent Template Induction with Gumbel-CRFs. Yao Fu, Chuanqi Tan, Alexander M. Rush, 2020

    https://arxiv.org/abs/2011.14244

    Html Tag Based Web Data Extraction and Tree Merging From Template Page

    Html Tag Based Web Data Extraction and Tree Merging From Template Page. A. Chandrasekhar, P. V. S. Readdy, 2014

    https://www.academia.edu/download/34193222/V2I3-0067.pdf

    Tree Automata

    https://github.com/topics/tree-automata

    What about trees and graphs?

    The original goal was to cover extensively the field of grammatical inference.

    This of course meant discussing in detail tree automata and grammars, giving the main adaptation of classical string algorithms to the case of trees, and even dealing with those works specific to trees. As work progressed it became clear that learning tree automata and grammars was going to involve at least as much material as with strings.

    The conclusion was reached to only sketch the specificities here, leaving the matter largely untouched, with everything to be written. This of course is not justified by the importance of the question, but only by the editorial difficulty and the necessity to stop somewhere. Of course, after trees will come the question of graphs...

    ...

    Extensions of the above mechanisms (automata, grammars) to deal with trees and graphs have been proposed. For the case of tree automata a general survey is (Comon et al., 1997) and for graph grammars there are a number of possible sources (Courcelle, 1991).

    ...

    Algorithm RPNI has been successfully adapted to tree automata (García & Oncina, 1993), and infinitary languages (de la Higuera & Janodet, 2004).

    ...

    In the field of computational linguistics, efforts have been made to learn context-free grammars from more informative data, such as trees (Charniak, 1996), following theoretical results by Yasubumi Sakakibara (Sakakibara, 1992). Learning from structured data has been a line followed by many: learning tree automata (Fernau, 2002, Habrard, Bernard & Jacquenet, 2002, Knuutila & Steinby, 1994), or context-free grammars from bracketed data (Sakakibara, 1990) allows to obtain better results, either with queries (Sakakibara, 1992), regular distributions (Carrasco, Oncina & Calera-Rubio, 2001, Kremer, 1997, Rico-Juan, Calera-Rubio & Carrasco, 2002), or negative information (García & Oncina, 1993). This has also led to different studies concerning the probability estimation of such grammars (Calera-Rubio & Carrasco, 1998, Lari & Young, 1990).

    ...

    19.3 About trees and graphs and more structure

    We have left untouched (or nearly untouched) the question of learning from data that would be more structured than strings. There are many researchers working on learning tree grammars and tree automata. In some cases the work consists of adapting a string language inference algorithm to suit the tree case, but in many others the problems are new and novel algorithms are needed. Furthermore, in practice, in many cases the tree structures allow us to model the data in a much more accurate fashion.

    ...

    19.5 About learning itself

    A view defended by some is that learning is about compressing; a compression with loss, where the loss itself corresponds to the gain in learning.

    Throughout the book we have viewed algorithms whose chief goal was to get hold of enormous amounts of data and somehow digest this into a simple set of rules which in turn allowed us to somehow replace the data by the grammar. In other words, the feeling we have reached is that learning is all about forgetting.

    — Colin de la Higuera - Grammatical Inference. Learning Automata and Grammars (2010)

    automata learning

    Substring-Based Algorithms

    Alignment-Based Learning

    author: van Zaanen 2000

    van Zaanen M (2000) ABL: alignment-based learning. In:Proceedings of the 18th international conference on computational linguistics (COLING), association for computational linguistics, association for computational linguistics, pp 961–967

    https://ilk.uvt.nl/menno/research/software/abl

    ABL learns structure from plain sequences (for example natural language sentences) by comparing them. Based on the parts of the sequences that are the same and parts that are not the same in two sequences, structure is inserted in the sequences.

    Grammatical Inference

    Error-Correcting Grammatical Inference

    author: Rulot and Vidal 1987

    Rulot H, Vidal E (1987) Modelling (sub)string-length based constraints through a grammatical inference method. In: Kittler J, Devijver P (eds) Proceedings of the NATO advanced study institute on pattern recognition theory and applications. Springer, pp 451–459

    ADIOS

    author: Solan et al. 2005

    Solan Z, Horn D, Ruppin E, Edelman S (2005) Unsupervised learning of natural languages. Proc Nat Acad Sci USA 102(33):11,629–11,634

    Data-Oriented Parsing

    author: Bod 2006

    Bod R (2006) An all-subtrees approach to unsupervised parsing. In: Proceedings of the 21st international conference on computational linguistics and 44th annual meeting of the ACL, association for computational linguistics, pp 865–872

    tree diff

    data transformers

    template-based parsing

    different from "template finding"

    aka: parser-generators

    javascript

    python

    java

    C++

    see also

    keywords

    • tree compression of many similar trees
    • reverse-engineer the template of many similar html files
    • find template from sample data
    • template-finding
    • template-detection
    • template-generation
    • template-generator
    • grammar generator
    • generate schema from data
    • approximate schema from data
    • find common structure of many similar html files
    • generate parser of many similar input files
    • reverse-engineering a JSX template from many rendered pages