Search & Query

Now & Then's Search, Query, and Browse tools are part of its powerful capabilities for finding, viewing, understanding information in context―and providing insight into your journal record.

Knowledge-based retrieval methodologies greatly increase the likelihood that you will find all and only (Recall & Precision) the journal information you want to review. Your information is retrieved based on its match to the concept and context of your query―you do not need to know (or match) any words or phrases in the information you wish to retrieve.

This extraordinary search capability is achieved using knowledge-based technologies to search the potentially large, but limited, domain of journal information. Your journal information has to do with only one person―you, and one view of the world―yours. Since the information is indexed, date parsed, grouped, filtered, and understood based only your worldview, the meaning and context of your queries is understood and ambiguities are resolved. Search methods include ad hoc, natural language queries, structured and prompted searches, and search templates.

Once the information that satisfies your query has been located, it is displayed and represented in formats that provide focus and insight (results lists, charts, narrative, timelines, hyperlinks, graphics, and mixed formats).

Word & Phrase Search
Everything you record in Now & Then is indexed for future access. At the most basic level, everything is searchable by key word or phrase. Key word searching implies that you know some words or phrases contained in the information you want to find. If, after one year, or 30 years you've forgotten important information, (forgotten) key word matching won't help―much more advanced retrieval techniques are needed.

Now & Then uses advanced techniques to relate the words, phrases, and content in your queries with your journal's stored content and concepts. Beyond full-text word and phrase matching, Now & Then employees word stemming, word proximity, misspelling and phonic reduction, synonyms, and user thesauruses. You can also use combinations of the Boolean connectors (AND and OR, NOT) with Date, Category, and Topic , and word proximity ranges to modify the reach and precision of your search.

Intelligent search
Although word-based search techniques may seem to successfully find the targeted information, they usually retrieve a lot of information that isn't relevant to your query. Phrase-based search methods also "miss" important and relevant information. To improve its ability to retrieve all and only the journal information you seek, Now & Then combines a number of advanced retrieval technologies that go beyond matching words and phrases―techniques that "understand" how you, individually, use language, and recognize your world, and your worldview.

Now & Then combines:
1) its understanding of the ways you individually use language with
2) its knowledge of your unique experiences, interests, thoughts, and philosophies (your worldview)

This enables Now & Then to recognize the content and context of your searches and queries and to gather retrieve precisely the relevant information while avoiding ambiguous and unrelated information. Now & Then monitors your journal's content, your queries, your use patterns, your feedback, personalization, and your personal rules to continuously improve its search and retrieval capabilities.

Limited Domain
Now & Then searches only journal information―it does not try to understand and search through huge stores of online information that have nothing to do with your life. As you use it, Now & Then learns about you, its user―a single, unique individual. It learns to understand the words and concepts you use to describe your activities, experiences and perceptions.

As Now & Then learns more about how you perceive the world, it makes increasingly better decisions about how your personal information should be stored, what information should be provided in response to your queries, and the information formats and presentation methods that provide crisp understanding and insight.

Now & Then's power and usefulness is continuously improved through machine learning, through your feedback, personalization, and training. Now & Then applies semantic indexing, multi-dimensional term scaling, classification, and concept clustering with Bayesian-based relevance feedback to understand your interests and worldview as they change throughout your life.

Over a lifetime, as you accumulate vast stores of journal information, the information remains you-centric. Although your interests, philosophies, personal and professional associations change dramatically, your journal information remains a special domain that can be accurately probed and explored based on Now & Then's accumulated knowledge about your evolving view of the world.
                                                                  
see also: language aware, domain aware  The Intelligent Journal:
Context Aware Search   ▪   Query interpretation technologies 

home