Information Retrieval III
Simhash - Procedure
1. Process the document into a set of features with associated weights. We will assume the simple case where the features are words weighted by their frequency. 2. Generate a hash value with b btis (the desired size of the fingerprint) for each word. The hash value shoudl be unique for each word. 3. In b-dimensional vertor V, update the components of the vector by adding the weight for a word to every component for which the corresponding bit in the words's hash value is 1, and subtracting the weight if the values is 0. 4. After all words have been processed, generate a b-bit fingerprint by setting the ith bit to 1 if the ith componed ov V is positive, or 0 otherwise.
Requirements for document storage system
1. Random access: request the content of a document based on its URL, hash function based on URL is typical. 2. Compression and large files: reducing storage requirements and efficient access. 3. Update: handling large volumes of new and modified documents, adding new anchor text
Near-Duplicate Detection - Tasks
1. Search: find near-duplicates of a document D, O(N) comparisons required. 2. Discovery: find all pairs of near-duplicate documents in the collection, O(N2) comparisons. 3. IR techniques are effective for search scenario. 4. For discovery, other techniques used to generate compact representations
Fingerprints
1. The document is parsed into words. Non-word content, such as punctuaion, HTML tags, and additional whitespace, is removed. 2. The words are grouped into contiguous n-grams for some n. These are usually overlapping sequences of words, although some techniques use non-overlapping sequences. 3. Some of the n-grams are selected to represent the document. 4. The selected n-grams are hashed to imporove retrieval efficiency and further reduce the size of the representation. 5. The hash values are stored, typically in an inverted index. 6. Documents are compared using overlap of finherprints.
Character Encoding definition
A character encoding is a mapping between bits and glyphs: i.e., getting from bits in a file to characters on a screen, Can be a major source of incompatibility. ASCII is basic character encoding scheme for English: encodes 128 letters, numbers, special characters, and control characters in 7 bits, extended with an extra bit for storage in bytes
Focused Crawling
Attempts to download only those pages that are about a particular topic: used by vertical search applications. Rely on the fact that pages about a topic tend to have links to other pages on the same topic: popular pages for a topic are typically used as seeds. Crawler uses text classifier to decide whether a page is on topic
columns, row keys
BigTable can have a huge number of ______ per row: all rows have the same column groups, not all rows have the same columns, important for reducing disk reads to access document data. Rows are partitioned into tablets based on their _______: simplifies determining which server is appropriate
Finding Content Blocks
Cumulative distribution of tags in the example web page: Main text content of the page corresponds to the "plateau" in the middle of the distribution
BigTable - Google's document storage system:
Customized for storing, finding, and updating web pages. Handles large collection sizes using inexpensive computers
Detecting Duplicates
Duplicate and near-duplicate documents occur in many situations: Copies, versions, plagiarism, spam, mirror sites, 30% of the web pages in a large crawl are exact or near duplicates of pages in the other 70%. Duplicates consume significant resources during crawling, indexing, and search: Little value to most users
Controlling Crawling
Even crawling a site slowly will anger some web server administrators, who object to any copying of their data. Robots.txt file can be used to control crawlers
Retrieving Web Pages
Every page has a unique uniform resource locator (URL). Web pages are stored on web servers that use HTTP to exchange information with client software
Duplicate Detection
Exact duplicate detection is relatively easy. Checksum techniques: A checksum is a value that is computed based on the content of the document, e.g., sum of the bytes in the document file; Possible for files with different text to have same checksum. Functions such as a cyclic redundancy check (CRC), have been developed that consider the positions of the bytes
Age
Expected age of a page t days after it was last crawled. Web page updates follow the Poisson distribution on average: Assume page updated on average once every lambda days, time until the next update is governed by an exponential distribution
Web Crawler
Finds and downloads web pages automatically: provides the collection for searching. Web is huge and constantly growing. Web is not under the control of search engine providers. Web pages are constantly changing. Crawlers also used for other types of data
HEAD, page changes
HTTP protocol has a special request type called _____ that makes it easy to check for ______
BigTable - Organization
Logically organized into rows. A row stores data for a single web page. Combination of a row key, a column key, and a timestamp point to a single cell in the row
UTF-32, UTF-8
Many applications use ______ for internal text encoding (fast random lookup) and _____ for disk storage (less space)
Document Feeds
Many documents are published: created at a fixed time and rarely updated again. e.g., news articles, blog posts, press releases, emails, tweets. Published documents from a single source can be ordered in a sequence called a document feed: new documents found by examining the end of the feed
Removing Noise
Many web pages contain text, links, and pictures that are not directly related to the main content of the page. This additional material is mostly noise that could negatively affect the ranking of the page. Techniques have been developed to detect the content blocks in a web page: Non-content material is either ignored or reduced in importance in the indexing process
Near-Duplicate Detection
More challenging task: Are web pages with same text context but different advertising or format near-duplicates?. A near-duplicate document is defined using a threshold value for some similarity measure between pairs of documents: e.g., document D1 is a near-duplicate of document D2 if more than 90% of the words in the documents are the same.
BigTable
No query language, no complex queries to optimize. Only row-level transactions. Tablets are stored in a replicated file system that is accessible by all BigTable servers. Any changes to a BigTable tablet are recorded to a transaction log, which is also stored in a shared file system. If any tablet server crashes, another server can immediately read the tablet data and transaction log from the file system and take over.
Freshness
Not possible to constantly check all pages: must check important pages and pages that change frequently. Freshness is the proportion of pages that are fresh. Optimizing for this metric can lead to bad decisions, such as not crawling popular sites (because of high cost of achieving freshness for them). Age is a better metric
Finding Content Blocks - Other Approaches
Other approaches use DOM structure and visual (layout) features
Character Encoding
Other languages can have many more glyphs: e.g., Chinese has more than 40,000 characters, with over 3,000 in common use. Many languages have multiple encoding schemes: e.g., CJK (Chinese-Japanese-Korean) family of East Asian languages, Hindi, Arabic, must specify encoding, can't have multiple languages in one file. Unicode developed to address encoding problems
Unicode
Proliferation of encodings comes from a need for compatibility and to save space: UTF-8 uses one byte for English (ASCII), as many as 4 bytes for some traditional Chinese characters, variable length encoding, more difficult to do string operations, UTF-32 uses 4 bytes for every character. Many applications use UTF-32 for internal text encoding (fast random lookup) and UTF-8 for disk storage (less space)
Document feed
Published documents from a single source can be ordered in a sequence called
RSS stands for
Really Simple Syndication, RDF Site Summary, Rich Site Summary
Finding Content Blocks - Optimeze, Maximize
Represent a web page as a sequence of bits, where bn = 1 indicates that the nth token is a tag. Optimization problem where we find values of i and j to maximize both the number of tags below i and above j and the number of non-tag tokens between i and j. i.e., maximize
Simhash
Similarity comparisons using word-based representations more effective at finding near-duplicates: Problem is efficiency. Simhash combines the advantages of the word-based similarity measures with the efficiency of fingerprints based on hashing. Similarity of two pages as measured by the cosine correlation measure is proportional to the number of bits that are the same in the simhash fingerprints.
Unicode Definition
Single mapping from numbers to glyphs that attempts to include all glyphs in common use in all known languages. Unicode is a mapping between numbers and glyphs: does not uniquely specify bits to glyph mapping!, e.g., UTF-8, UTF-16, UTF-32.
Sitemaps
Sitemaps contain lists of URLs and data about those URLs, such as modification time and modification frequency. Generated by web server administrators. Tells crawler about pages it might not otherwise find. Gives crawler a hint about when to check a page for changes
Deep Web
Sites that are difficult for a crawler to find are collectively referred to as the deep (or hidden) Web: much larger than conventional Web. Three broad categories: 1. private sites: no incoming links, or may require log in with a valid account. 2. form results: sites that can be reached only after entering some data into a form. 3. scripted pages: pages that use JavaScript, Flash, or another client-side language to generate links
Web Crawler Procedure Seeds
Starts with a set of seeds, which are a set of URLs given to it as parameters. Seeds are added to a URL request queue. Crawler starts fetching pages from the request queue. Downloaded pages are parsed to find link tags that might contain other useful URLs to fetch. New URLs added to the crawler's request queue, or frontier. Continue until no more new URLs or disk full
large files, overhead
Store many documents in _______, rather than each document in a file: avoids ________ in opening and closing files, reduces seek time relative to read time
Compression
Text is highly redundant (or predictable). Compression techniques exploit this redundancy to make files smaller without losing any of the content. Compression of indexes covered later. Popular algorithms can compress HTML and XML text by 80%: e.g., DEFLATE (zip, gzip) and LZW (UNIX compress, PDF), may compress large files in blocks to make access faster
Conversion
Text is stored in hundreds of incompatible file formats: e.g., raw text, RTF, HTML, XML, Microsoft Word, ODF, PDF. Other types of files also important: e.g., PowerPoint, Excel. Typically use a conversion tool: converts the document content into a tagged text format such as HTML or XML, retains some of the important formatting information
costs, crawl
The older a page gets, the more it _____ not to _____ it: e.g., expected age with mean change frequency λ = 1/7 (one change per week) (unit of time is one day)
Distributed Crawling
Three reasons to use multiple computers for crawling: 1. Helps to put the crawler closer to the sites it crawls. 2. Reduces the number of sites the crawler has to remember. 3. Reduces computing resources required. Distributed crawler uses a hash function to assign URLs to crawling computers: hash function should be computed on the host part of each URL
Document Feeds - Documen Types
Two types: A push feed alerts the subscriber to new documents, A pull feed requires the subscriber to check periodically for new documents. Most common format for pull feeds is called RSS: Really Simple Syndication, RDF Site Summary, Rich Site Summary, or ...
Desktop Crawls
Used for desktop search and enterprise search. Differences to web crawling: 1. Much easier to find the data. 2. Responding quickly to updates is more important. 3. Must be conservative in terms of disk and CPU usage. 4. Many different document formats. 5. Data privacy very important (multiple user accounts)
Retrieving Web Pages Procedure DNS
Web crawler client program connects to a domain name system (DNS) server. DNS server translates the hostname into an internet protocol (IP) address. Crawler then attempts to connect to server host using specific port. After connection, crawler sends an HTTP request to the web server to request a page usually a GET request
Web Crawling - Reduce efficiency
Web crawlers spend a lot of time waiting for responses to requests. To reduce this inefficiency, web crawlers use threads and fetch hundreds of pages at once. Crawlers could potentially flood sites with requests for pages. To avoid this problem, web crawlers use politeness policies: e.g., delay between requests to same web server
Freshness - Web Pages
Web pages are constantly being added, deleted, and modified. Web crawler must continually revisit pages it has already crawled to see if they have changed in order to maintain the freshness of the document collection: stale copies no longer reflect the real contents of the web pages
Database systems
can provide document storage for some applications: web search engines use customized document storage systems
Unicode Examples
e.g., Greek letter pi (π) is Unicode symbol number 960, In binary, 00000011 11000000 (3C0 in hexadecimal), Final encoding is 11001111 10000000 (CF80 in hexadecimal)
Reasons to store converted document text
saves crawling time when page is not updated. Provides efficient access to text for snippet generation, information extraction, etc.
RSS
ttl tag (time to live): amount of time (in minutes) contents should be cached, In practice: how often a feed should be refreshed. RSS feeds are accessed like web pages: using HTTP GET requests to web servers that host them. Easy for crawlers to parse. Easy to find new information.
Compound documents formats
used to store multiple documents in a file, e.g., TREC Web