Artificial Intelligence Paula Matuszek

©2006 Paula Matuszek What is Artificial Intelligence l Definitions –The science and engineering of making intelligent machines, especially intelligent computer programs. It is related to the similar task of using computers to understand human intelligence, but AI does not have to confine itself to methods that are biologically observable. (McCarthy, 2002) –The exciting new effort to make computers think... machines with minds, in the full and literal sense (Haugeland, 1985) –The automation of activities that we associate with human thinking, activities such as decision-making, problem solving, learning... (Bellman, 1978) l Strong AI and Weak AI l Turing Test

©2006 Paula Matuszek What Methods Does AI Use? l AI can also be defined in terms of what kinds of methods it uses –Search –Knowledge Representation –Inference –Logic –Pattern recognition –Machine Learning

©2006 Paula Matuszek Typical AI Domains l Games l Natural Language Processing l Planning l Perception l Robotics l Expert Systems l Intelligent Agents

©2006 Paula Matuszek So when WILL we decide that computers are intelligent?

©2006 Paula Matuszek How Do We Know When We're There? l Some requirements I think any test we use must meet: –Whatever test we use must not exclude the majority of adult humans. I can't play chess at a grand master level! –Whatever test we use must produce an observable result. "Isn't intelligent because it doesn't have a mind" is perhaps a topic for interesting philosophical debate, but it's not of any practical help.

©2006 Paula Matuszek What can AI systems do? Here are some example applications l Computer vision: face recognition from a large set l Robotics: autonomous (mostly) car l Natural language processing: simple machine translation l Expert systems: medical diagnosis in narrow domain l Spoken language systems: ~1000 word continuous speech l Planning and scheduling: Hubble Telescope experiments l Learning: text categorization into ~1000 topics l User modeling: Bayesian reasoning in Windows help l Games: Grand Master level in chess (world champion), checkers, etc.

©2006 Paula Matuszek What can’t AI systems do yet? l Understand natural language robustly (e.g., read and understand articles in a newspaper) l Surf the web l Interpret an arbitrary visual scene l Learn a natural language l Play Go well l Construct plans in dynamic real-time domains l Refocus attention in complex environments l Perform life-long learning

©2006 Paula Matuszek AI Uses in Information Science l Retrieval l Ontologies l Intelligent Agents l Text Mining

©2006 Paula Matuszek Challenges and Possibilities l Information overload. There’s too much. We would like –Better retrieval –Help with handling documents we have –Help finding specific pieces of information without having to read documents l What might help? –Statistical techniques –Natural language processing techniques –Knowledge domain based techniques

©2006 Paula Matuszek Retrieval l Find correct documents, with high precision and high recall. l AI used extensively for: –Determining relevance: heuristic rules capture human intuition about importance. Improves precision –Using domain models: using domain models/ontologies with synonyms and classes improves recall.

©2006 Paula Matuszek Retrieval: Some Current Directions l Intelligent spiders –Can't cover all of the web; it's too big! –Determine relevance as documents are retrieved; spider only those with high relevance –Goal is to improve precision AND recall l Intelligent disambiguation –When you search for "bank" do you mean the financial institution or the side of a river? –Use ontologies to find multiple meanings –Scan for related words to choose meaning l Semantic web –Add meta-information as you create web pages. Intelligent data instead of intelligent tools.

©2006 Paula Matuszek Ontologies l Definition: An ontology is a formal description or specification of the concepts and relationships in a domain. l Synonyms, hierarchy of terms, richer relations. l Example: cat –Synonyms: pussy, feline, kitty –Is a: mammal, pet –Subclass: Persian, Siamese, tabby –Has characteristics: carnivorous, purrs

©2006 Paula Matuszek Ontology: Another Example l Example: Panadol –Broader term: chemical drug substance –Narrower term: acetaminophen tablet –synonyms: Tylenol, acetaminophen, paracetamol –Preferred term: paracetamol –Trademarked in country: UK, US, EU. –Company-holding-trademark: SmithKline –Ingredient-in: Contac –USAN: acetaminophen –BAN: paracetamol –Therapeutic class: analgesic agent, antipyretic agent

©2006 Paula Matuszek Intelligent Agents l Definition: a software program which autonomously gathers information or performs some task for a user. l Communicative l Capable l Autonomous l Adaptive

©2006 Paula Matuszek Some Current Intelligent Agent Tasks l Screen out junk mail –Understand what makes mail junk: Hand-built rules or machine learning l Shopbots: Find the best price for X –Know about and access shopping sites –Know about and understand costing: Price for items, discounts, shipping fees l News and mail alerts –Understand what I am interested in –Watch relevant sources to find those things and bring them to my attention l Recommender systems –What movies or books might I be interested in? –Collaborative systems, faceted or characteristic- based systems.

©2006 Paula Matuszek Intelligent Agents: The Vision l Lucy calls her brother Pete: "Mom needs to see a specialist and then has to have a series of physical therapy sessions. I'm going to have my agent set up the appointments." Pete agrees to share driving. l At the MD office, Lucy instructs her agent through her handheld browser. The agent –retrieves information about Mom's prescribed treatment from the doctor's agent –looks up several lists of providers –checks for the ones in-plan for Mom's insurance within a 20-mile radius of her home and with a rating of excellent or very good on trusted rating services –finds a match between available appointment times (supplied by the agents of individual providers through their Web sites) and Pete's and Lucy's busy schedules. l The agent presents a plan. Pete doesn't like it – too much driving, and at rush hour, and has his agent redo the search with stricter preferences about location and time. Lucy's agent, having complete trust in Pete's agent in the context of the present task, supplies the data it has already sorted through. l A new plan is presented: a closer clinic and earlier times—with warning notes. –Pete will have to reschedule a couple of his less important appointments. –The insurance company's list does not include this provider under physical therapists: "Service type and insurance plan status securely verified by other means. (Details?)" l Lucy and Pete agree and the agent makes the appointments. l Pete asks his agent to explain how it had found that provider even though it wasn't on the proper list. Example taken from Scientific American article on the Semantic Web, May,

©2006 Paula Matuszek Text Mining l Common theme: information exists, but in unstructured text. l Text mining is the general term for a set of techniques for analyzing unstructured text in order to process it better –Document-based –Content-based

©2006 Paula Matuszek Document-Based l Techniques which are concerned with documents as a whole, rather than details of the contents –Document retrieval: find documents –Document categorization: sort documents into known groups –Document classification: cluster documents into similar classes which are not predefined –Visualization: visually display relationships among documents

©2006 Paula Matuszek Document Categorization l Document categorization –Assign documents to pre-defined categories l Examples –Process into work, personal, junk –Process documents from a newsgroup into “interesting”, “not interesting”, “spam and flames” –Process transcripts of bugged phone calls into “relevant” and “irrelevant” l Issues –Real-time? –How many categories/document? Flat or hierarchical? –Categories defined automatically or by hand?

©2006 Paula Matuszek Categorization -- Automatic l Statistical approaches similar to search engine l Set of “training” documents define categories –Underlying representation of document is bag of words (BOW): looking at frequencies, not at order –Category description is created using neural nets, regression trees, other Machine Learning techniques –Individual documents categorized by net, inferred rules l Requires relatively little effort to create categories l Accuracy is heavily dependent on "good" training examples l Typically limited to flat, mutually exclusive categories

©2006 Paula Matuszek Categorization: Manual l Natural Language/linguistic techniques l Categories are defined by people –underlying representation of document is stream of tokens –category description contains –ontology of terms and relations –pattern-matching rules –individual documents categorized by pattern-matching l Defining categories can be very time-consuming l Typically takes some experimentation to "get it right" l Can handle much more complex structures

©2006 Paula Matuszek Document Classification l Document classification –Cluster documents based on similarity l Examples –Group samples of writing in an attempt to determine author(s) –Look for “hot spots” in customer feedback –Find new trends in a document collection (outliers, hard to classify) l Getting into areas where we don’t know ahead of time what we will have; true “mining”

©2006 Paula Matuszek Document Classification -- How l Typical process is: –Describe each document –Assess similarities among documents –Establish classification scheme which creates optimal "separation" l One typical approach: –document is represented as term vector –cosine similarity for measuring association –bottom-up pairwise combining of documents to get clusters l Assumes you have the corpus in hand

©2006 Paula Matuszek Document Clustering l Approaches vary a great deal in –document characteristics used to describe document (linguistic or semantic? bow? –methods used to define "similar" –methods used to create clusters l Other relevant factors –Number of clusters to extract is variable –Often combined with visualization tools based on similarity and/or clusters –Sometimes important that approach be incremental l Useful approach when you don't have a handle on the domain or it's changing

©2006 Paula Matuszek Document Visualization l Visualization –Visually display relationships among documents l Examples –hyperbolic viewer based on document similarity; browse a field of scientific documents –“map” based techniques showing peaks, valleys, outliers –Faceted search results showing document counts for different categorizations, with browsing l Highly interactive, intended to aid a human in finding interrelationships and new knowledge in the document set.

©2006 Paula Matuszek Content-Based Text Mining l Methods which focus in a specific document rather than a corpus of documents –Document Summarization: summarize document –Feature Extraction: find specific features –Information Extraction: find detailed information l Often not interested in document itself

©2006 Paula Matuszek Document Summarization l Document Summarization –Provide meaningful summary for each document l Examples: –Search tool returns “context” –Monthly progress reports from multiple projects –Summaries of news articles on the human genome l Often part of a document retrieval system, to enable user judge documents better l Surprisingly hard to make sophisticated

©2006 Paula Matuszek Document Summarization -- How l T wo general approaches: –Extract representative sentences/clauses: extractive –Capture document in generic representation and generate summary: abstractive l Extractive –If in response to search, keywords. Easy, effective –Otherwise term frequency, position, etc; Broadly applicable, gets "general feel“. Current state of art. l Abstractive –Create "template" or "frame" –NL processing to fill in frame –Generation based on template Good if well-defined domain, clearcut information needs. Hard.

©2006 Paula Matuszek Feature Extraction l Group individual terms into more complex entities (which then become tokens) l Examples –Dates, times, names, places –URLs, HREFs and IMG tags – Relationships like “X is president of Y” l Can involve quite high-level features: language l Enables more sophisticated queries –Show me all the people mentioned in the news today –Show me every mention of “New York” l Also refers to extracting aspects of document which somehow characterize it: length, vocab, etc

©2006 Paula Matuszek Information Extraction l Retrieve some specific information which is located somewhere in this set of documents. l Don’t want the document itself, just the info. –Information may occur multiple times in many documents, but we just need to find it once –Often what is really wanted from a web search. l Tools not typically designed to be interactive; not fast enough for interactive processing of a large number of documents l Often first step in creating a more structured representation of the information

©2006 Paula Matuszek Some Examples of Information Extraction l Financial Information –Who is the CEO/CTO of a company? –What were the dividend payments for stocks I’m interested in for the last five years? l Biological Information –Are there known inhibitors of enzymes in a pathway? –Are there chromosomally located point mutations that result in a described phenotype? l Other typical questions –who is familiar with or working on a domain? –what patent information is available?

©2006 Paula Matuszek l Create a model of information to be extracted l Create knowledge base of rules for extraction –concepts –relations among concepts l Find information –Word-matching: template. "Open door". –Shallow parsing: simple syntax. "Open door with key" –Deep Parsing: produce parse tree from document l Process information (into database, for instance) l Involves some level of domain modeling and natural language processing Information Extraction -- How

©2006 Paula Matuszek Why Text Is Hard Natural language processing is AI-Complete. l Abstract concepts are difficult to represent l LOTS of possible relationships among concepts l Many ways to represent similar concepts l Tens or hundreds or thousands of features/dimensions

©2006 Paula Matuszek Text is Hard l I saw Pathfinder on Mars with a telescope. l Pathfinder photographed Mars. l The Pathfinder photograph mars our perception of a lifeless planet. l The Pathfinder photograph from Ford has arrived. l The Pathfinder forded the river without marring its paint job.

©2006 Paula Matuszek Why Text is Easy l Highly redundant when you have a lot of it l Many relatively crude methods provide fairly good results: –Pull out “important” phrases –Find “meaningfully” related words –Create summary from document –“grep” l Evaluating results is not easy; need to know the question!